The 2025 lung cancer landscape: advances in screening, molecular taxonomy and therapeutic strategy: a narrative review

Introduction

Lung cancer remains a leading cause of cancer mortality worldwide (1). Although cigarette smoking has declined in many regions, lung cancer epidemiology is undergoing a structural transition (2). An increasing proportion of cases occurs in individuals with little or no smoking exposure, and risk is increasingly shaped by convergent environmental, occupational, and metabolic inflammatory determinants (3). In parallel, the clinical landscape is being transformed by two complementary forces. First, early detection is expanding beyond narrowly defined high-risk groups, bringing feasibility (4), overdiagnosis, capacity constraints, and implementation science to the centre of debate. Second, therapeutic innovation spanning perioperative immunotherapy and targeted approaches, ultra-sensitive circulating tumor DNA (ctDNA) assays, antibody-drug conjugates (ADCs), and next-generation immunotherapies has improved outcomes in selected populations (5,6), while introducing new complexities related to toxicity, sequencing, affordability, and equitable access. Accordingly, this review addresses three pragmatic questions: (I) which shifts in risk stratification and screening strategy are most likely to change population-level outcomes; (II) which biological advances in 2025 are closest to translational or clinical leverage (biomarkers or therapeutic vulnerabilities); and (III) how emerging diagnostics and treatments should be interpreted in the context of evidence maturity, toxicity, feasibility, and equity.

In this review, we synthesize major peer-reviewed advances reported in 2025 with an emphasis on translational and clinical impact (Figure 1). We take a continuum-of-care perspective, integrating progress in (I) risk stratification and health-system constraints; (II) biological mechanisms of initiation, progression and metastasis; (III) screening and precision diagnostics; (IV) perioperative and locally advanced management; (V) systemic therapy for advanced non-small cell lung cancer (NSCLC); (VI) small cell lung cancer (SCLC) and other neuroendocrine (NE) lung malignancies; and (VII) survivorship, health economics and global delivery. Citations follow the numbering scheme used throughout the article to enable direct cross-referencing across sections. Throughout, we move beyond reporting effect sizes to critically appraise limitations, generalisability, and implementation constraints that shape real-world impact. We present this article in accordance with the Narrative Review reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1477/rc).

Figure 1 Conceptual map of major advances in lung cancer research and clinical management in 2025. Figure concept and layout designed by the authors; initial artwork generated using Nano Banana Pro and subsequently edited by the authors for scientific accuracy and clarity. ADC, antibody drug conjugate; cfDNA, cell-free DNA; CNS, central nervous system; CT, computed tomography; ctDNA, circulating tumour DNA; DLL3, delta-like ligand 3; EGFR, epidermal growth factor receptor; LCINS, lung cancer in never-smokers; LDCT, low-dose computed tomography; MET, mesenchymal epithelial transition factor; MRD, minimal residual disease; NSCLC, non-small cell lung cancer; OS, overall survival; PD-1, programmed cell death 1; PFS, progression-free survival; SCLC, small cell lung cancer; TKI, tyrosine kinase inhibitor; VEGF, vascular endothelial growth factor.

Methods

This narrative review synthesises peer-reviewed advances in lung cancer research and clinical management published primarily in 2025 and organises them along a continuum-of-care framework spanning: (I) shifting risk structures and health-system constraints; (II) biological mechanisms of tumour initiation, progression and metastasis; (III) screening and precision diagnostics; (IV) perioperative and locally advanced management; (V) systemic therapy for advanced NSCLC; (VI) SCLC and other pulmonary NE malignancies; and (VII) survivorship, health economics and global delivery. Our objective was to provide an evidence-informed, implementation-aware synthesis of developments most likely to influence clinical decision-making, translational prioritisation, or policy-relevant delivery across this pathway.

Literature identification, search strategy, and time window

Literature was identified through structured database searches and targeted surveillance of high-impact journals. We searched PubMed/MEDLINE, Embase, Web of Science and complemented this with targeted surveillance of high-impact general and specialty journals that frequently publish pivotal thoracic oncology research (including N Engl J Med, The Lancet/Lancet Oncology, JAMA, BMJ, Nature/Science/Cell, and major clinical/translational counterparts). We also screened major oncology meeting presentations/abstracts [American Society of Clinical Oncology (ASCO), European Society for Medical Oncology (ESMO), and World Conference on Lung Cancer (WCLC)] when relevant to capture practice-defining trial updates. The prespecified primary timeframe was 1 January 2025 to 18 December 2025, with limited extension to late 2024 or early 2026 when needed to capture essential trial context (e.g., background standards of care, mature follow-up updates, or companion analyses that materially affected interpretation). Reference lists of key papers were screened to identify closely linked supportive, contrasting, or contextual studies, including negative or inconclusive results relevant to clinical interpretation.

Search terms were mapped to the major sections of the review and combined disease terms with theme-specific keywords. Major concept blocks included: screening/early detection [low-dose computed tomography (LDCT), non-risk-based screening, risk model, eligibility, overdiagnosis, nodule management, adherence, navigation, implementation]; diagnostics (navigational bronchoscopy, transthoracic biopsy, diagnostic yield, complication, pneumothorax); computational stratification (digital pathology, artificial intelligence (AI), virtual molecular testing, tumour cellularity, calibration, bias); liquid biopsy [ctDNA, cell-free DNA (cfDNA) fragmentomics, ultrasensitive detection, minimal residual disease (MRD), perioperative kinetics, molecular staging]; perioperative/locally advanced therapy (neoadjuvant, adjuvant, chemo-immunotherapy, pathological response, event-free survival, overall survival, consolidation, timing); metastatic NSCLC [antibody-drug conjugate, bispecific antibody, sequencing, resistance, interstitial lung disease (ILD), pneumonitis]; and SCLC/NE (neuronal coupling, synapse-like programmes, DLL3, T-cell engager, radiotherapy, brain metastases). To mitigate selective emphasis on positive findings, we also incorporated terms capturing null/negative outcomes and harm signals (e.g., “no improvement”, “failed”, “negative trial”, “discontinuation”, “toxicity”, “ILD”, “pneumonitis”) within major screening and therapeutic domains. A summary of the search strategy is provided in Table 1.

Eligibility and selection principles

Because this is a narrative synthesis, we did not aim to enumerate all studies published in 2025. Instead, we applied explicit selection principles to define “influential” evidence as research meeting at least one of the following criteria: (I) practice-informing clinical evidence, prioritising phase III randomised trials, guideline-shaping studies, and platform studies that plausibly change perioperative, locally advanced, metastatic, or SCLC treatment algorithms; (II) programme-level screening or diagnostic evidence from high-quality prospective cohorts, pragmatic trials, or modelling analyses with clear implications for eligibility, nodule-management thresholds, harms/overdiagnosis, and deliverability; (III) mechanistic and translational studies that advance causal understanding of carcinogenesis, tumour evolution, metastasis, or neuro-immune-metabolic coupling, supported by multi-omics (single-cell/spatial profiling, lineage tracing, or evolutionary inference) together with functional validation; and (IV) implementation relevance, including tissue constraints, turnaround time, infrastructure requirements, toxicity monitoring, access, affordability, and equity.

Exclusion criteria were: non-peer-reviewed items without sufficient methodological detail; studies without clear relevance to lung cancer clinical management or mechanistic understanding; and highly preliminary signals lacking validation when higher-quality evidence existed for the same question. When multiple studies addressed the same topic, we prioritised those with stronger design, broader generalisability, more mature follow-up, or clearer clinical implications.

Selection process

Titles/abstracts retrieved from the database search were screened by two reviewers independently W.F. and X. Zou). Potentially eligible records underwent full-text assessment by the same reviewers. Disagreements were resolved through discussion and consensus; if consensus could not be reached, a senior reviewer W.L. adjudicated. Studies identified through targeted journal surveillance and reference-list screening were assessed using the same eligibility principles and selection process.

Global landscape and macro-level challenges: shifting risk structures and rising health-system pressure

Lung cancer control is increasingly defined by a transition from a single dominant exposure model to cumulative, multi-factorial risk accrued over decades. As screening programmes broaden and precision therapies proliferate, the most consequential constraints are often those of implementation, including service capacity, affordability, longitudinal engagement, and the ability to deliver complex pathways equitably at scale.

In a comprehensive overview of the United States (U.S.) cancer statistics for 2025, Siegel et al. (7) reaffirmed that lung cancer remains among the leading causes of cancer death, while emphasizing that the composition of attributable risk is changing. Although cigarette smoking still contributes the largest fraction, an increasing share of risk is distributed across non-cigarette exposures, including secondhand smoke, other combustible tobacco products, radon, occupational hazards, and ambient air pollution. Complementing this epidemiological framing, Islami et al. (8) quantified the long-term impact of tobacco control in the U.S. from 1970 to 2022, estimating 3,856,240 lung cancer deaths averted and 76,275,550 life-years gained. Together, these analyses sharpen a dual imperative: sustain and strengthen tobacco control, while modernizing risk assessment beyond smoking history alone towards integrated, exposure-informed stratification.

Beyond established carcinogenic exposures, epidemiological evidence in 2025 further implicated lifestyle-associated determinants and systemic host biology as modifiers of lung cancer risk. Using the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial cohort, Wang et al. (9) reported an association between higher ultra-processed food consumption and increased lung cancer incidence [highest versus lowest quartile: overall hazard ratio (HR) =1.41; NSCLC HR =1.37; SCLC HR =1.44]. As with observational nutritional epidemiology, residual confounding and measurement error remain important caveats; nonetheless, these results align with a broader emerging narrative in which metabolic and inflammatory perturbations may modulate carcinogenic susceptibility and tumour evolution, motivating deeper mechanistic and interventional work.

As early detection transitions from efficacy to implementation at scale, real-world effectiveness is increasingly determined by programme delivery rather than test performance alone. Key attrition points include eligibility ascertainment, appointment completion, longitudinal adherence, and timely linkage to diagnostic work-up and treatment. In a subgroup analysis of the pragmatic INHALE randomized trial, Baggett et al. (10) reported that patient navigation increased LDCT completion among individuals with current or prior homelessness (current homelessness: 26.8% versus 7.1%; prior homelessness: 51.3% versus 10.2%). However, the more modest absolute gains in the currently unhoused group indicate that navigation, while beneficial, does not fully offset structural barriers. These findings support an implementation agenda that couples navigation with system-level adaptations, including low-threshold access models, co-located or mobile services, flexible scheduling, sustained case management, and financing approaches that reduce out-of-pocket and opportunity costs. Together, the data suggest that further advances in lung cancer early detection will depend as much on delivery architecture and equity-focused policy as on continued technological refinement.

The biological backdrop of tumour initiation and progression: clonal dynamics, neuro microenvironment coupling, and systemic niches

Two biological directions were particularly prominent in 2025. First, experimental and computational advances enabled more traceable and quantitative accounts of early carcinogenesis. Second, lung tumours were increasingly conceptualised as ecosystem processes shaped by coupled neural, immune, and metabolic circuits, with implications for progression, dissemination, and therapy resistance.

Using a carcinogen-induced lung squamous cell carcinoma model, Gómez-López et al. (11) showed that carcinogen exposure induces non-neutral competition among airway basal cells, driving abnormal clonal expansion and basal cell mobilisation along the airway epithelium. The resulting field effects led to multifocal premalignant lesions that were frequently dominated by a small number of highly mutated clones occupying extensive regions of the bronchial tree, consistent with a discrete shift in clonal fitness during field cancerization. Multi-region sequencing in human samples provided convergent support for clonally related premalignant lesions across spatially separated airway regions.

For lung cancer in never-smokers, Díaz-Gay et al. (12) analysed 871 tumours from the multinational Sherlock-Lung cohort and reported substantial geographic and exposure-associated heterogeneity in mutational processes. SBS40a, of currently unknown aetiology, contributed prominently in adenocarcinoma and was enriched in epidermal growth factor receptor (EGFR)-mutant tumours, whereas SBS22a, consistent with aristolochic acid exposure, was observed in Taiwan. Notably, secondhand smoke exposure was not associated with a distinct driver spectrum or mutational signature, whereas higher air pollution exposure correlated with increased TP53 mutations, telomere shortening, and a higher mutation burden with dose-response patterns, including increases in SBS4 and SBS5.

At the level of evolutionary reconstruction, Pawlik et al. (13) introduced allele-specific phylogenetic analysis of copy number alterations (ALPACA) to improve joint inference of single-nucleotide variants and somatic copy-number alterations. In the TRACERx421 cohort, they linked dissemination-associated clones to increased chromosomal instability, and showed that greater clonal copy-number diversity was associated with worse disease-free survival, positioning copy-number heterogeneity as a quantifiable prognostic dimension.

In 2025, SCLC biology moved beyond correlative descriptions towards experimentally tractable causal models of tumour neuron interaction. Peinado et al. (14) reported that NE SCLC cells, but not non-NE cells, are electrically excitable, and that action potential firing directly promotes malignant phenotypes. The energetic costs of excitability imposed a high ATP demand and an unexpected dependence on oxidative phosphorylation in NE cells, whereas non-NE cells provided metabolic support in a manner reminiscent of neuron glia coupling. Disease progression was accompanied by remodelling of the innervation landscape, consistent with a self-reinforcing loop linking electrical activity, metabolic dependency, and neural remodelling.

Sakthivelu et al. (15) provided convergent structural and functional evidence that SCLC establishes synapse-like contacts with neurons and receives neurotransmission. Electrophysiological recordings and optogenetic perturbations supported neuronal inputs mediated through NMDA and GABA_A receptors. Co-culture with vagal sensory or cortical neurons enhanced proliferative capacity, and pharmacological inhibition of glutamatergic signalling yielded therapeutic signals in spontaneous mouse models, positioning synaptic signalling co-option as a potentially actionable vulnerability. At the organismal level, Savchuk et al. (16) showed that neuronal activity-dependent mechanisms extend beyond local interactions. Vagotomy suppressed primary lung tumour initiation and progression, whereas glutamatergic and GABAergic activity promoted brain tumour proliferation through paracrine and synaptic routes. Notably, tumour cell membrane depolarisation and calcium transients were sufficient to enhance intracranial growth. Together, these studies establish a testable framework for neuro coupled SCLC and motivate therapeutic exploration of excitability programmes, calcium signalling, glutamate and GABA pathways, and peripheral innervation.

Beyond tumour-derived mediators that condition distant organs, metabolites within the target organ microenvironment can act as instructive cues for metastatic adaptation. Doglioni et al. (17) reported elevated aspartate in lung interstitial fluid in breast cancer patients and in mouse models, and showed that extracellular aspartate activates an NMDAR CREB signalling axis in tumour cells, induces DOHH, and promotes eIF5A dependent non-canonical translation programmes, with transforming growth factor beta signalling and collagen synthesis forming a central effector hub that enhances lung metastatic invasiveness. This work delineates a mechanistic sequence linking nutrient cues to receptor-mediated transcriptional and translational reprogramming, and highlights multiple points that may be therapeutically tractable.

At the systemic level, metabolic stress intersects with coagulation and innate immune pathways to shape pre-metastatic permissiveness. Wu et al. (18) showed that glucose restriction or low-carbohydrate conditions suppressed primary tumours yet promoted lung metastasis via exosomal TRAIL, with downstream polarisation of poliovirus receptor (PVR) positive macrophages and induction of natural killer (NK) cell exhaustion; TIGIT blockade attenuated metastatic risk while enhancing anti-tumour effects. Lucotti et al. (19) described a pro-thrombotic lung niche in which CXCL13 reprogrammed stromal macrophages released extracellular vesicles bearing integrin beta 2, which engaged platelet GPIb and triggered aggregation, thereby promoting cancer-associated thrombosis and metastasis. Together, these studies connect metabolism, innate immunity, and coagulation through defined pathways that yield candidate biomarkers and druggable targets.

Therapeutic targeting in 2025 increasingly focused on vulnerabilities defined by in vivo dependency rather than in vitro essentiality, particularly those linked to chromatin state control and resistance to regulated cell death. Jeong et al. (20) described a clinical-grade catalytic inhibitor of NSD2 and showed that pharmacological disruption of the NSD2 H3K36me2 axis remodelled aberrant chromatin plasticity, suppressed tumour growth, and prolonged survival in KRAS-driven lung and pancreatic cancer models. The reported synergy with the KRAS inhibitor sotorasib further supports a rationale for combination strategies that couple oncogene blockade with epigenetic state modulation.

Complementing this theme, Wu et al. (21) provided evidence that lung adenocarcinoma exhibits a pronounced in vivo reliance on ferroptosis suppression that is not fully captured by in vitro assays. Genetic loss of FSP1 increased lipid peroxidation and markedly impaired tumour initiation and growth, and pharmacological FSP1 inhibition conferred therapeutic benefit across models, positioning FSP1 as a translationally relevant ferroptosis target. In parallel, Schneider et al. (22) identified nucleotide metabolism, specifically GUK1, as an actionable metabolic liability in lung cancer, extending the repertoire of druggable dependencies beyond canonical signalling nodes.

Mechanistic studies of lung immune homeostasis and tissue repair reported in 2025 provide useful reference frameworks for interpreting tumour-associated inflammation and treatment-related lung injury. Hoagland et al. (23) showed that macrophage-derived oncostatin M mitigated type I interferon mediated suppression of epithelial regeneration, promoted ATII proliferation, and supported organoid formation in models of inflammatory lung damage. Deng et al. (24) further delineated the tissue and cell-type specificity of MR1 expression, reporting particularly high levels in lung and peritoneal macrophages. In this context, MR1-dependent antigen presentation shaped MAIT-cell immunity and influenced lung microbiota composition, supporting a myeloid MR1 MAIT axis as a regulator of mucosal immune homeostasis.

As immunotherapy expands into perioperative NSCLC, differences in efficacy increasingly demand a structured account of tumour immune microenvironment states. Liu et al. (25) generated integrated single-cell and spatial transcriptomic profiles of resected NSCLC specimens following anti programmed cell death protein 1 (PD-1) treatment and proposed a tumour immune microenvironment classification framework. Major pathological response was linked to defined immune composition and activation features, including immature CXCL13 positive B cells, interferon gamma positive helper T cells, and FGFBP2 positive NK cells. By contrast, non-major pathological response tumours were enriched for immunosuppressive populations such as CCR8 positive regulatory T cells and exhibited stronger inhibitory signalling, including higher CD47. These data provide operational candidates for response prediction, rational stratification, and mechanism-informed combination design in the perioperative setting.

Screening and precision diagnosis: from population strategy to digital and multimodal stratification

Screening and diagnostic research in 2025 advanced along two convergent lines. One focused on broadening eligibility beyond traditional high-risk definitions anchored in smoking history. The other aimed to shift diagnosis from morphology-led decision-making towards multimodal, computationally supported stratification.

Critical appraisal of screening expansion in 2025. Across non-risk-based or broadened-risk screening reports, a recurring limitation is reliance on surrogate endpoints (detection rate and stage shift) rather than demonstrated mortality benefit, while the magnitude of overdiagnosis remains uncertain—particularly for indolent ground-glass-predominant lesions. In parallel, expanded eligibility can amplify downstream diagnostic burden (repeat imaging, invasive biopsy, complication risk) and strain system capacity (radiology workload, nodule clinics, navigation, and timely linkage to treatment). Therefore, policy-relevant screening thresholds and nodule-management rules should be calibrated not only to baseline risk and imaging distributions but also to delivery feasibility, adherence, and equity; prospective implementation studies that jointly measure benefits, harms, and capacity constraints are essential.

In a prospective community cohort, Li et al. (26) evaluated non risk based LDCT screening in adults aged 40 to 74 years and reported that lung cancer detection in guideline-defined non-high-risk groups still reached 1.6% under National Comprehensive Cancer Network (NCCN) criteria and 1.3% under the Chinese consensus, with stage I accounting for 92.0% and 93.2% of detected cancers, respectively. By inference, a substantial proportion of cancers in this screened population would have been missed under smoking-centred eligibility thresholds, with estimated miss rates of 81.0% using NCCN criteria and 44.0% using the Chinese consensus. In the U.S., Bandi et al. (27) used modelling to estimate deaths prevented and life-years gained under full implementation of LDCT screening, reinforcing that population benefit is strongly contingent on delivery, uptake, and adherence. In the UK SUMMIT programme, Bhamani et al. (28) reported high sensitivity (97.0%) and specificity (95.2%) for LDCT in a diverse high-risk population, supporting operational feasibility at scale.

Evidence from large screening datasets also informed recalibration of nodule management parameters. Using extensive Chinese screening data, Ye et al. (29) proposed increasing the malignancy risk threshold applied to pure ground-glass nodules from 5 to 8 mm, with the goal of improving specificity and reducing unnecessary surveillance and overdiagnosis pressure. More broadly, this work reinforces the view that nodule-management thresholds are not fixed technical constants but policy-relevant parameters that should be calibrated to baseline risk, imaging distributions, and health-system capacity.

For diagnostic work-up of intermediate-to-high-risk peripheral nodules (10 to 30 mm), Lentz et al. (30) conducted a multicentre noninferiority trial comparing navigational bronchoscopy with CT-guided transthoracic needle biopsy. Navigational bronchoscopy achieved noninferior diagnostic accuracy at 12-month confirmation (79.0% versus 73.6% for transthoracic biopsy) while substantially reducing pneumothorax (3.3% versus 28.3%) and downstream interventions such as chest tube placement and hospitalisation. This trial provides pragmatic evidence to guide pathway selection based on a balance between diagnostic yield and procedure-related morbidity.

Computational approaches expanded the practical scope of precision diagnosis when tissue, turnaround time, or infrastructure are limiting. In this setting, AI-based tools can support triage and prioritisation—for example, by flagging slides more likely to harbour actionable drivers or by estimating tumour cellularity to reduce failed molecular tests. Importantly, these models should be positioned as decision-support systems rather than replacements for molecular assays, and their value depends on prospective validation within real workflows, population calibration, and transparent reporting of misclassification risks. Zhao et al. (31) reported DeepGEM, a model that predicts key driver mutations directly from routine hematoxylin and eosin (H&E) slides, achieving areas under the curve (AUCs) of 0.90 to 0.97 in internal testing and 0.80 to 0.91 in external multicentre validation, with generalisation to lymph-node metastases (EGFR AUC 0.91; KRAS AUC 0.88). This supports a role for digital pathology as triage and decision support rather than a replacement for molecular assays. Wang et al. (32) described DECIPHER-NODL, which integrates LDCT imaging features with cfDNA fragmentomics through stacked ensembling, achieving high discrimination for benign versus malignant nodules (internal AUC 0.950; external AUC 0.966) and improved specificity at fixed sensitivity. They also reported an invasiveness prediction model with AUC of approximately 0.88 to inform management intensity.

Ultra-sensitive ctDNA technologies progressed from recurrence surveillance towards molecular staging and risk stratification in early disease. In the TRACERx cohort, Black et al. (33) reported a tumor-informed whole-genome ctDNA platform with sensitivity down to 1 to 3 parts per million and specificity of 99.9%, enabling preoperative ctDNA detection in 81% of lung adenocarcinoma cases, including 53% of stage I tumors. Importantly, even very low-level preoperative ctDNA positivity below 80 parts per million was associated with worse overall survival, supporting the concept that molecular detection thresholds can capture clinically relevant risk states not visible to conventional staging. In a companion longitudinal analysis, Black et al. (5) tracked up to approximately 1,800 patient-specific variants across perioperative and surveillance time points, showing that ctDNA presence and kinetics provide high-resolution prognostic information and may refine inference on relapse timing and patterns.

Early-stage and locally advanced disease: the perioperative era anchored by long-term overall survival and ctDNA dynamics

A defining shift in perioperative research in 2025 was a move from short-term response metrics towards validation on overall survival, alongside the emergence of ctDNA dynamics as a practical substrate for risk-adaptive escalation and de-escalation. This maturation of endpoints reflects both the need to confirm that improvements in pathological response and event-free survival translate into durable benefit, and the growing availability of biomarkers capable of refining postoperative decision-making.

Forde et al. (34) reported 5-year follow-up from CheckMate 816 in resectable stage IB to IIIA NSCLC, showing that three cycles of neoadjuvant nivolumab plus chemotherapy improved overall survival compared with chemotherapy alone (HR =0.72; P=0.048), with 5-year overall survival of 65.4% versus 55.0% at a median follow-up of 68.4 months. Exploratory analyses also strengthened the link between pathological complete response and long-term outcome, with 5-year overall survival of 95.3% among patients achieving pathological complete response compared with 55.7% among those without pathological complete response, supporting pathological complete response as a clinically meaningful intermediate endpoint in this setting.

Biomarker-informed tailoring of adjuvant intensity also advanced. Zhao et al. (35) reported that among patients achieving major pathological response or pathological complete response after induction immunotherapy, adjuvant immunotherapy was not associated with additional survival benefit, raising the possibility that a subset of patients may be overtreated and highlighting the rationale for prospective de-escalation approaches. In parallel, Spigel et al. (36) reported AIM-HIGH, a randomized phase III trial in stage IA to IIA non-squamous NSCLC defined as molecular high-risk by a clinical laboratory improvement amendments (CLIA)-certified 14-gene expression profile, in which platinum-based adjuvant chemotherapy improved disease-free survival at interim analysis (24-month disease-free survival 96% versus 79%; HR =0.22; P=0.0087). Together, these studies illustrate complementary roles for dynamic response biomarkers and static risk classifiers in postoperative decision-making, with implications for matching treatment intensity to residual risk.

Evidence for perioperative targeted therapy in oncogene-driven NSCLC also strengthened. In NeoADAURA, He et al. (37) reported that in resectable EGFR-mutated stage II to IIIB NSCLC, neoadjuvant osimertinib plus chemotherapy, and osimertinib monotherapy, increased major pathological response rates to 26% and 25%, respectively, compared with 2% with chemotherapy alone, without new perioperative safety signals. For ALK-positive potentially resectable stage III disease, Leonetti et al. (38) reported the final analysis of the ALNEO phase II trial, supporting feasibility and meeting prespecified pathological endpoints for perioperative alectinib. Collectively, these results extend the perioperative paradigm to molecularly defined subgroups and accelerate the evidence chain for targeted approaches around surgery.

Perioperative regimen development continued to broaden the therapeutic ceiling while testing the boundaries of feasibility. Awad et al. (39) compared neoadjuvant nivolumab plus ipilimumab without chemotherapy with chemotherapy in resectable lung cancer and reported a signal of benefit, although the comparison was exploratory and confidence intervals crossed unity. Cascone et al. (40) reported the platform phase II NeoCOAST-2 trial, adding oleclumab, monalizumab, or the TROP2 antibody drug conjugate datopotamab deruxtecan to durvalumab plus chemotherapy, with pathological complete response rates of 20.3%, 25.7%, and 35.2% across arms and major pathological response rates of 41.9%, 50.0%, and 63.0%, alongside surgical completion exceeding 93%, supporting rapid perioperative iteration with preserved operability. In parallel, surgical risk stratification and operative decision-making remained salient. Sun et al. (41) reported that respiratory sarcopenia stratified the risk of severe postoperative complications in NSCLC. Li et al. (42) showed that selective pneumonectomy after neoadjuvant therapy with or without immunotherapy can be feasible, providing data to inform risk benefit trade-offs for complex resections in the immunotherapy era.

For locally advanced disease, 2025 provided both consolidation signals and timing cautions. In limited-stage SCLC, Zenke et al. (43) reported a Japanese exploratory analysis from the phase III ADRIATIC trial supporting overall survival and progression-free survival benefits with durvalumab consolidation after concurrent chemoradiotherapy, with manageable safety. By contrast, in unresectable stage III NSCLC, Bradley et al. (44) reported PACIFIC-2, in which initiating durvalumab concurrently from the start of chemoradiotherapy and continuing as consolidation did not improve progression-free survival (HR =0.85; P=0.247) or overall survival (HR =1.03; P=0.823), and was associated with higher discontinuation due to adverse events. These results indicate that earlier immunotherapy is not inherently superior, and that timing, tolerability, and completion remain central determinants of net benefit in stage III disease.

Systemic therapy for advanced NSCLC: treatment reconfiguration driven by antibody drug conjugates and bispecific antibodies

A defining clinical development in 2025 was the maturation of ADCs and bispecific antibodies into phase III validated options across molecularly and histologically defined subgroups. Across multiple lines of therapy, decision-making began to shift away from a default chemotherapy backbone towards biologically stratified strategies that combine payload delivery with immune and vascular co-targeting.

Early-phase data positioned TROP2-directed ADCs as clinically active in advanced NSCLC. Zhao et al. (45) reported meaningful single-agent activity of sacituzumab tirumotecan (sac TMT), with an overall objective response rate of 40% and a median progression-free survival of 6.2 months; post hoc analyses suggested higher activity in EGFR-mutant disease, with an objective response rate of 55% and a median progression-free survival of 11.1 months. Fang et al. (46) extended this signal into phase III in EGFR tyrosine kinase inhibitor (TKI) resistant, EGFR-mutated non-squamous NSCLC, showing that sac TMT improved progression-free survival compared with pemetrexed plus platinum (8.3 versus 4.3 months; HR =0.49) and was associated with an overall survival benefit (HR =0.60). In a separate randomized study, Fang et al. (47) showed that sac TMT outperformed docetaxel in previously treated EGFR-mutated advanced NSCLC (objective response rate 45% versus 16%; progression-free survival 6.9 versus 2.8 months; HR =0.30), with favourable 12-month survival signals. Together, these trials reposition TROP2 ADCs as central options in the post TKI setting for EGFR-mutant disease.

Evidence for additional ADC platforms also strengthened. In the phase III TROPION Lung01 trial, Ahn et al. (6) reported that datopotamab deruxtecan improved progression-free survival compared with docetaxel (HR =0.75) but did not significantly improve overall survival, with the benefit largely driven by non-squamous histology. In the phase II TROPION Lung05 study, Sands et al. (48) reported sustained activity of datopotamab deruxtecan in genomically actionable subgroups following targeted therapy and platinum chemotherapy. Beyond TROP2, Steuer et al. (49) reported activity of patritumab deruxtecan (HER3 DXd) after platinum and immunotherapy. Li et al. (50) reported high response rates with trastuzumab rezetecan in HER2-mutant NSCLC with a manageable safety profile, while reinforcing ILD and pneumonitis as a toxicity domain requiring systematic monitoring.

Bispecific antibodies also produced practice-relevant phase III signals, particularly for combined immune and vascular targeting. In programmed death-ligand 1 (PD-L1) positive advanced NSCLC without sensitising EGFR mutations or ALK fusions, Xiong et al. (51) reported the phase III HARMONi 2 trial, showing that the PD-1 and vascular endothelial growth factor (VEGF) bispecific antibody ivonescimab improved progression-free survival compared with pembrolizumab (11.1 versus 5.8 months; HR =0.51), with consistent benefit in both tumor proportion score (TPS) 1% to 49% and TPS at least 50% subgroups. In first-line advanced squamous NSCLC, Chen et al. (52) reported HARMONi 6, in which ivonescimab plus chemotherapy improved progression-free survival compared with tislelizumab plus chemotherapy (11.1 versus 6.9 months; HR =0.60) and the benefit was independent of PD-L1 expression. Collectively, these results position dual immune and vascular blockade as a competitive benchmark for first-line therapy.

Targeted therapy evidence density also increased across common and rarer genomic subtypes. For EGFR-mutated advanced NSCLC, Yang et al. (53) reported final overall survival results from MARIPOSA, showing that amivantamab plus lazertinib improved overall survival compared with osimertinib, with consistent benefit in higher-risk subgroups including patients with brain metastases. Jänne et al. (54) reported overall survival benefit in FLAURA2 with osimertinib plus platinum and pemetrexed compared with osimertinib alone, supporting a stratified approach in which treatment intensity is weighed against toxicity and patient risk. For EGFR exon 20 insertion disease, Yang et al. (55) reported WU KONG1B, showing objective response rates of approximately 46% to 47% after platinum therapy with dose-dependent trade-offs, including higher rates of grade at least 3 diarrhoea at higher dose levels. Piotrowska et al. (56) reported activity of zipalertinib in platinum-pretreated patients with or without prior amivantamab exposure, including central nervous system activity. For HER2-mutant NSCLC, Le et al. (57) reported sevabertinib with objective response rates ranging from 38% to 71% across cohorts, and Heymach et al. (58) reported Beamion LUNG-1, in which the selective irreversible HER2 inhibitor zongertinib achieved a confirmed objective response rate of 71%, median duration of response of 14.1 months, and median progression-free survival of 12.4 months, with low rates of grade at least 3 adverse events and no drug-related ILD signal.

For KRAS G12C, studies further refined sequencing and combination strategies. Barlesi et al. (59) reported KRYSTAL 12, showing that adagrasib improved progression-free survival compared with docetaxel (HR =0.58). Sacher et al. (60) reported longer follow-up for divarasib, supporting continued interest in higher potency inhibitors. In early combination data, Zhong et al. (61) reported glecirasib plus an SHP2 inhibitor with high objective response rates in treatment-naïve patients but limited activity after prior KRAS G12C inhibition, underscoring persistent resistance constraints. For ROS1 and BRAF V600E, Pérol et al. (62) reported high objective response rates with taletrectinib in treatment-naïve ROS1-positive NSCLC, and Johnson et al. (63) updated overall survival analyses supporting encorafenib plus binimetinib in BRAF V600E-mutant NSCLC.

Finally, evidence continued to accumulate for integrating local therapy with systemic treatment in selected advanced settings. Zhou et al. (64) reported 4-year outcomes from GEMSTONE 302, confirming sustained survival benefit with sugemalimab plus chemotherapy in the first-line setting. In EGFR-mutated oligometastatic NSCLC, Sun et al. (65) reported a randomized phase III trial showing that adding concurrent thoracic radiotherapy (60 Gy) to first-line EGFR TKI improved both progression-free survival and overall survival (median progression-free survival 17.1 versus 10.6 months; HR =0.57; median overall survival 34.4 versus 26.2 months; HR =0.62), with a modest increase in severe treatment-related adverse events. These findings support a role for local consolidation strategies to deliver incremental survival gains in biologically selected populations even in the setting of effective targeted therapy.

Critical appraisal of the 2025 metastatic NSCLC therapeutics evidence. First, overall survival remains immature for multiple ADC and bispecific programmes, and cross-trial comparisons are confounded by rapidly evolving standards of care and heterogeneous post-progression access. Second, toxicity is a central determinant of real-world adoption: ILD/pneumonitis, ocular toxicity, and myelosuppression can be dose-limiting and may disproportionately affect patients with pre-existing lung disease or prior radiotherapy. Third, many trials enrol selected populations with biomarker enrichment and intensive monitoring, which may limit generalisability and complicate sequencing decisions in routine practice. Finally, costs and infrastructure requirements (biomarker testing, toxicity surveillance, and supportive care) create implementation and equity constraints that must be weighed alongside efficacy.

SCLC and pulmonary NE neoplasms: immune redirection and precision radiotherapy in parallel

A major translational advance in 2025 for SCLC was the clinical validation of immune redirection strategies targeting DLL3. Mountzios et al. (66) reported the phase III DeLLphi 304 trial, showing that the DLL3 by CD3 bispecific T cell engager tarlatamab improved overall survival compared with investigator’s choice chemotherapy in SCLC after platinum therapy (13.6 versus 8.3 months; HR =0.60; P<0.001), with accompanying improvements in progression-free survival. Patient-reported outcomes also favoured tarlatamab, including dyspnoea and cough, and the rate of grade at least 3 adverse events was lower than with chemotherapy (54% versus 80%). Extending the development trajectory into earlier treatment phases, Paulson et al. (67) reported DeLLphi 303, evaluating tarlatamab plus a PD-L1 inhibitor as first-line maintenance following chemo-immunotherapy in extensive-stage SCLC. At a median follow-up of approximately 18.4 months, the median overall survival was 25.3 months, providing a rationale for ongoing phase III evaluation of combination strategies in the maintenance setting.

In extensive-stage SCLC, maintenance intensification and radiotherapy refinement progressed in parallel, with an increasing focus on balancing survival benefit against functional outcomes and cumulative toxicity. Paz-Ares et al. (68) reported IMforte, in which patients without progression after induction atezolizumab plus carboplatin and etoposide were randomized to maintenance lurbinectedin plus atezolizumab versus atezolizumab alone. The combination improved progression-free survival (HR =0.54; P<0.0001) and overall survival (HR =0.73; P=0.017), but increased grade 3 to 4 adverse events (38% versus 22%) and highlighted clinically relevant myelosuppression, including a higher proportion of grade 5 events. For prophylactic cranial irradiation, Gondi et al. (69) reported NRG CC003, showing that hippocampal avoidance prophylactic cranial irradiation (PCI) was noninferior to standard PCI for 12-month intracranial relapse (14.7% versus 14.8%). Although it did not significantly improve the primary cognitive endpoint, defined as 6-month decline in Hopkins Verbal Learning Test-Revised (HVLT-R) delayed recall, it reduced the risk of any neurocognitive test failure and yielded similar overall survival and grade at least 3 toxicity. Aizer et al. (70) reported a prospective phase II strategy for SCLC patients with 1 to 10 brain metastases under strict imaging surveillance, in which stereotactic radiosurgery or radiotherapy achieved a 1-year neurological death rate of 11.0% and only 22% of patients required salvage whole-brain radiotherapy, supporting a surveillance-intensive stereotactic approach in selected patients to reduce or defer whole-brain radiotherapy exposure.

Beyond SCLC, pulmonary NE neoplasms remain an area of increasing clinical relevance with limited prospective evidence. Hallet et al. (71) analysed contemporary incidence and survival trends and reported a rising incidence and poor outcomes for large-cell NE carcinoma, highlighting persistent gaps in prospective trials and the need for molecularly informed stratification to support reproducible treatment pathways.

Health economics and global strategy: value-based full-cycle governance

Health-system considerations featured more prominently in 2025, reflecting the widening gap between what is biologically feasible and what can be delivered sustainably at scale. Work in this area increasingly treated value, access, and institutional capacity as core components of the lung cancer continuum rather than peripheral implementation issues.

Tan et al. (72) quantified survival outcomes and medication costs across biomarker-defined subgroups in NSCLC and reported that patients with actionable driver alterations experienced longer survival and lower medication cost per survivor than biomarker-negative patients, suggesting that improving access and affordability for biomarker-negative populations remains a priority for payers and drug development. Moving from individual-level burden towards system-level design, Pramesh et al. (73) summarised the International Association for the Study of Lung Cancer (IASLC) Lancet Commission and proposed a set of transformative initiatives for global lung cancer control over the coming decade, including affordable screening, equitable molecular testing, and stigma reduction. The commission emphasised that durable progress will depend not only on technology diffusion but also on governance and institutional capacity building across health systems.

As survival improves, special populations and late effects are increasingly shaping the agenda of “routine” care. Rivera et al. (74) compared screening patterns in cancer survivors with those in individuals without a history of cancer and suggested that screening strategies may require re-evaluation and greater standardisation in survivor populations. In parallel, Ma et al. (75) conducted a systematic review and network meta-analysis of cardiovascular adverse events associated with EGFR tyrosine kinase inhibitors in EGFR-mutated NSCLC. They reported increased cardiac risk compared with placebo for both first- and third-generation EGFR TKIs, with higher overall risk for third-generation agents and a more pronounced arrhythmia signal, underscoring the need to embed toxicity assessment and monitoring earlier within baseline evaluation and longitudinal follow-up frameworks.

Discussion

The 2025 evidence base suggests that lung cancer care is entering a full-lifecycle phase in which population risk, early detection, biological taxonomy, and treatment sequencing co-evolve. However, progress is heterogeneous in evidence maturity and deliverability. Screening expansion beyond smoking-centred eligibility aligns with shifting epidemiology, yet it is constrained by uncertain overdiagnosis magnitude, variable adherence, and system capacity. Similarly, multimodal diagnostics—including AI-enabled pathology and imaging-cfDNA classifiers—expand feasibility when tissue is limited, but clinical value depends on prospective validation within representative workflows and on transparent management of failure modes.

Biological advances provide a bridge to these clinical shifts. Field cancerization and copy-number-driven fitness help explain why early lesions can be spatially distributed and evolutionarily constrained, supporting the rationale for risk models that incorporate non-smoking exposures and for biomarkers that capture clonal dynamics. In SCLC, convergent data on neuronal coupling and synapse-like programmes move the field beyond association towards mechanistic tractability, motivating exploration of excitability programmes, calcium signalling, and neurotransmission pathways as therapeutic vulnerabilities.

Therapeutically, the maturation of ADCs and bispecific antibodies is reshaping metastatic sequencing, but it also foregrounds deliverability: toxicity monitoring, pulmonary safety adjudication, and affordability are now integral to “clinical benefit”. In perioperative disease, long-term overall survival strengthens the neoadjuvant chemo-immunotherapy paradigm, while ctDNA kinetics offers a practical substrate for response-adaptive escalation and de-escalation—yet harmonisation of assays, thresholds, and pre-analytical pipelines is required before broad adoption.

Collectively, the central challenge is to integrate discovery and delivery. Future trials and programmes should embed pragmatic endpoints (implementation, adherence, toxicity management, and equity) alongside efficacy, and should standardise biomarker frameworks (especially ctDNA) to support interoperable decision-making across the care continuum.

Conclusions

The 2025 evidence base indicates that lung cancer management is entering a full-lifecycle phase characterised by dynamic stratification, systems integration, and explicit attention to value. Epidemiologically, the rising burden of lung cancer in never-smokers and the prominence of multi-exposure risk argue for a transition from smoking-centred risk models towards more inclusive frameworks that incorporate environmental and metabolic inflammatory dimensions. Biologically, lineage tracing, evolutionary reconstruction, and single-cell and spatial atlases have clarified key steps in early carcinogenesis and have advanced neuro coupled SCLC into experimentally tractable territory, sharpening hypotheses about where intervention may be most effective. Clinically, screening strategies are expanding beyond conventional high-risk gating, while AI-enabled pathology, multimodal imaging and cfDNA models, and ultra-sensitive ctDNA assays are extending precision diagnosis and molecular risk stratification towards higher throughput and earlier timepoints. In resectable disease, long-term overall survival follow-up consolidates neoadjuvant chemo-immunotherapy as a benchmark approach, and ctDNA kinetics provides a plausible foundation for response-adaptive escalation and de-escalation trials. In advanced NSCLC, ADCs and bispecific antibodies are reshaping sequencing while bringing new priorities to the foreground, including optimal patient selection, toxicity management, affordability, and implementation across diverse healthcare settings. Actionable priorities for research and policy. First, broadened-risk or non-risk-based screening should be evaluated in pragmatic designs that jointly quantify benefits (mortality or robust proxies), harms (overdiagnosis, procedure complications), and system capacity (workload, nodule pathways, linkage-to-care). Second, nodule-management thresholds should be treated as policy parameters and calibrated to population risk and delivery feasibility. Third, AI-enabled pathology and imaging-cfDNA classifiers require prospective, workflow-embedded validation, population calibration, and governance frameworks that define accountability and mitigate bias. Fourth, for perioperative NSCLC, ctDNA-guided escalation/de-escalation trials should prioritise assay harmonisation, interpretable thresholds, and clinically usable turnaround. Fifth, for metastatic NSCLC, sequencing studies should explicitly incorporate pulmonary safety adjudication (ILD/pneumonitis), real-world generalisability, and affordability to ensure that efficacy translates into equitable outcome gains. Future progress will increasingly depend on integrating molecular and ecological tumour biology with pragmatic delivery science and health policy, to ensure that innovation translates into durable and equitable outcome gains.

Acknowledgments

We acknowledge the use of Nano Banana Pro (an AI-based image generation tool) to generate the initial draft artwork for Figure 1. The figure concept, scientific content, layout refinement, and final rendering were developed and substantially edited by the authors to ensure accuracy and clarity.

Footnote

Reporting Checklist: The authors have completed the Narrative Review reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1477/rc

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1477/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1-1477/coif). W.L. serves as an unpaid Associate Editor-in-Chief of Translational Lung Cancer Research from May 2025 to April 2026. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. Luo G, Zhang Y, Rumgay H, et al. Estimated worldwide variation and trends in incidence of lung cancer by histological subtype in 2022 and over time: a population-based study. Lancet Respir Med 2025;13:348-63. [Crossref] [PubMed]
3. Naidu SB, Wisking A, Karoshi A, et al. Risk Factors Associated With Incidence of Lung Cancer in Never-Smokers: A Systematic Review and Meta-Analysis. JTO Clin Res Rep 2025;6:100910. [Crossref] [PubMed]
4. Silvestri GA, Young RP, Tanner NT, et al. Screening Low-Risk Individuals for Lung Cancer: The Need May Be Present, but the Evidence of Benefit Is Not. J Thorac Oncol 2024;19:1155-63. [Crossref] [PubMed]
5. Black JRM, Karasaki T, Abbott CW, et al. Longitudinal ultrasensitive ctDNA monitoring for high-resolution lung cancer risk prediction. Cell 2025;188:7083-7098.e18. [Crossref] [PubMed]
6. Ahn MJ, Tanaka K, Paz-Ares L, et al. Datopotamab Deruxtecan Versus Docetaxel for Previously Treated Advanced or Metastatic Non-Small Cell Lung Cancer: The Randomized, Open-Label Phase III TROPION-Lung01 Study. J Clin Oncol 2025;43:260-72. [Crossref] [PubMed]
7. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
8. Islami F, Nargis N, Liu Q, et al. Averted lung cancer deaths due to reductions in cigarette smoking in the United States, 1970-2022. CA Cancer J Clin 2025;75:216-25. [Crossref] [PubMed]
9. Wang K, Zhao J, Yang D, et al. Association between ultra-processed food consumption and lung cancer risk: a population-based cohort study. Thorax 2025;80:810-8. [Crossref] [PubMed]
10. Baggett TP, Sporn N, Barbosa Teixeira J, et al. Homelessness, Patient Navigation, and Lung Cancer Screening in a Health Center Setting: A Subgroup Analysis of a Randomized Clinical Trial. JAMA Netw Open 2025;8:e2519780. [Crossref] [PubMed]
11. Gómez-López S, Alhendi ASN, Przybilla MJ, et al. Aberrant basal cell clonal dynamics shape early lung carcinogenesis. Science 2025;388:eads9145. [Crossref] [PubMed]
12. Díaz-Gay M, Zhang T, Hoang PH, et al. The mutagenic forces shaping the genomes of lung cancer in never smokers. Nature 2025;644:133-44. [Crossref] [PubMed]
13. Pawlik P, Grigoriadis K, Bunkum A, et al. Clone copy number diversity is linked to survival in lung cancer. Nature 2025;646:190-7. [Crossref] [PubMed]
14. Peinado P, Stazi M, Ballabio C, et al. Intrinsic electrical activity drives small-cell lung cancer progression. Nature 2025;639:765-75. [Crossref] [PubMed]
15. Sakthivelu V, Schmitt A, Odenthal F, et al. Functional synapses between neurons and small cell lung cancer. Nature 2025;646:1243-53. [Crossref] [PubMed]
16. Savchuk S, Gentry KM, Wang W, et al. Neuronal activity-dependent mechanisms of small cell lung cancer pathogenesis. Nature 2025;646:1232-42. [Crossref] [PubMed]
17. Doglioni G, Fernández-García J, Igelmann S, et al. Aspartate signalling drives lung metastasis via alternative translation. Nature 2025;638:244-50. [Crossref] [PubMed]
18. Wu CY, Huang CX, Lao XM, et al. Glucose restriction shapes pre-metastatic innate immune landscapes in the lung through exosomal TRAIL. Cell 2025;188:5701-5716.e19. [Crossref] [PubMed]
19. Lucotti S, Ogitani Y, Kenific CM, et al. Extracellular vesicles from the lung pro-thrombotic niche drive cancer-associated thrombosis and metastasis via integrin beta 2. Cell 2025;188:1642-1661.e24. [Crossref] [PubMed]
20. Jeong J, Hausmann S, Dong H, et al. NSD2 inhibitors rewire chromatin to treat lung and pancreatic cancers. Nature 2026;649:205-15. [Crossref] [PubMed]
21. Wu K, Vaughan AJ, Bossowski JP, et al. Targeting FSP1 triggers ferroptosis in lung cancer. Nature 2026;649:487-95. [Crossref] [PubMed]
22. Schneider JL, Kurmi K, Dai Y, et al. GUK1 activation is a metabolic liability in lung cancer. Cell 2025;188:1248-1264.e23. [Crossref] [PubMed]
23. Hoagland DA, Rodríguez-Morales P, Mann AO, et al. Macrophage-derived oncostatin M repairs the lung epithelial barrier during inflammatory damage. Science 2025;389:169-75. [Crossref] [PubMed]
24. Deng J, Yan Y, Zhang X, et al. Macrophage MR1 antigen presentation promotes MAIT cell immunity and lung microbiota modulation. Science 2026;391:eadr6322. [Crossref] [PubMed]
25. Liu Z, Yang Z, Wu J, et al. A single-cell atlas reveals immune heterogeneity in anti-PD-1-treated non-small cell lung cancer. Cell 2025;188:3081-3096.e19. [Crossref] [PubMed]
26. Li C, Cheng B, Li J, et al. Non-Risk-Based Lung Cancer Screening With Low-Dose Computed Tomography. JAMA 2025;333:2108-10. [Crossref] [PubMed]
27. Bandi P, Landy R, Star J, et al. Lung Cancer Deaths Prevented and Life-Years Gained From Lung Cancer Screening. JAMA 2025;334:2225-7. [Crossref] [PubMed]
28. Bhamani A, Creamer A, Verghese P, et al. Low-dose CT for lung cancer screening in a high-risk population (SUMMIT): a prospective, longitudinal cohort study. Lancet Oncol 2025;26:609-19. [Crossref] [PubMed]
29. Ye W, Fu W, Li C, et al. Diameter thresholds for pure ground-glass pulmonary nodules at low-dose CT screening: Chinese experience. Thorax 2025;80:76-85. [Crossref] [PubMed]
30. Lentz RJ, Frederick-Dyer K, Planz VB, et al. Navigational Bronchoscopy or Transthoracic Needle Biopsy for Lung Nodules. N Engl J Med 2025;392:2100-12. [Crossref] [PubMed]
31. Zhao Y, Xiong S, Ren Q, et al. Deep learning using histological images for gene mutation prediction in lung cancer: a multicentre retrospective study. Lancet Oncol 2025;26:136-46. [Crossref] [PubMed]
32. Wang H, Huang H, Li F, et al. Risk-stratified classification of pulmonary nodule malignancy via a machine learning model integrating imaging and cell-free DNA: a model development and validation study (DECIPHER-NODL). Lancet Reg Health West Pac 2025;64:101730. [Crossref] [PubMed]
33. Black JRM, Bartha G, Abbott CW, et al. Ultrasensitive ctDNA detection for preoperative disease stratification in early-stage lung adenocarcinoma. Nat Med 2025;31:70-6. [Crossref] [PubMed]
34. Forde PM, Spicer JD, Provencio M, et al. Overall Survival with Neoadjuvant Nivolumab plus Chemotherapy in Lung Cancer. N Engl J Med 2025;393:741-52. [Crossref] [PubMed]
35. Zhao ZR, Yan WP, Yu XY, et al. Adjuvant immunotherapy does not improve survival in non-small cell lung cancer with major/complete pathologic response after induction immunotherapy. J Thorac Cardiovasc Surg 2025;169:1576-1584.e3. [Crossref] [PubMed]
36. Spigel DR, Westeel V, Anderson IC, et al. Adjuvant chemotherapy for stage IA-IIA non-squamous, non-small-cell lung cancer identified as molecular high-risk by a 14-gene expression profile (AIM-HIGH): an international, randomised, phase 3 trial. Lancet Respir Med 2025;13:887-96. [Crossref] [PubMed]
37. He J, Tsuboi M, Weder W, et al. Neoadjuvant Osimertinib for Resectable EGFR-Mutated Non-Small Cell Lung Cancer. J Clin Oncol 2025;43:2875-87. [Crossref] [PubMed]
38. Leonetti A, Boni L, Gnetti L, et al. Alectinib as neoadjuvant treatment in potentially resectable stage III ALK-positive NSCLC: Final analysis of ALNEO phase II trial (GOIRC-01-2020-ML42316). J Clin Oncol 2025;43:8015.
39. Awad MM, Forde PM, Girard N, et al. Neoadjuvant Nivolumab Plus Ipilimumab Versus Chemotherapy in Resectable Lung Cancer. J Clin Oncol 2025;43:1453-62. [Crossref] [PubMed]
40. Cascone T, Bonanno L, Guisier F, et al. Perioperative durvalumab plus chemotherapy plus new agents for resectable non-small-cell lung cancer: the platform phase 2 NeoCOAST-2 trial. Nat Med 2025;31:2788-96. [Crossref] [PubMed]
41. Sun C, Hirata Y, Kawahara T, et al. Diagnosis of Respiratory Sarcopenia for Stratifying Postoperative Risk in Non-Small Cell Lung Cancer. JAMA Surg 2025;160:66-73. [Crossref] [PubMed]
42. Li Z, Wu L, Wang C, et al. Outcomes After Neoadjuvant Therapy With or Without Immunotherapy Followed By Pneumonectomy in Non-Small Cell Lung Cancer Patients. Ann Thorac Surg 2025;120:118-28. [Crossref] [PubMed]
43. Zenke Y, Shiraishi Y, Goto Y, et al. Durvalumab Post Concurrent Chemoradiotherapy in Japanese Patients With Limited-Stage Small-Cell Lung Cancer in the Phase 3 ADRIATIC Trial. Cancer Sci 2025;116:3171-84. [Crossref] [PubMed]
44. Bradley JD, Sugawara S, Lee KH, et al. Simultaneous Durvalumab and Platinum-Based Chemoradiotherapy in Unresectable Stage III Non-Small Cell Lung Cancer: The Phase III PACIFIC-2 Study. J Clin Oncol 2025;43:3610-21. [Crossref] [PubMed]
45. Zhao S, Cheng Y, Wang Q, et al. Sacituzumab tirumotecan in advanced non-small-cell lung cancer with or without EGFR mutations: phase 1/2 and phase 2 trials. Nat Med 2025;31:1976-86. [Crossref] [PubMed]
46. Fang W, Wu L, Meng X, et al. Sacituzumab Tirumotecan in EGFR-TKI-Resistant, EGFR-Mutated Advanced NSCLC. N Engl J Med 2026;394:13-26. [Crossref] [PubMed]
47. Fang W, Li X, Wang Q, et al. Sacituzumab tirumotecan versus docetaxel for previously treated EGFR-mutated advanced non-small cell lung cancer: multicentre, open label, randomised controlled trial. BMJ 2025;389:e085680. [Crossref] [PubMed]
48. Sands J, Ahn MJ, Lisberg A, et al. Datopotamab Deruxtecan in Advanced or Metastatic Non-Small Cell Lung Cancer With Actionable Genomic Alterations: Results From the Phase II TROPION-Lung05 Study. J Clin Oncol 2025;43:1254-65. [Crossref] [PubMed]
49. Steuer CE, Hayashi H, Su WC, et al. Patritumab Deruxtecan (HER3-DXd; MK-1022) in Non-Small Cell Lung Cancer After Platinum-Based Chemotherapy and Immunotherapy. J Clin Oncol 2025;43:2816-26. [Crossref] [PubMed]
50. Li Z, Wang Y, Sun Y, et al. Trastuzumab rezetecan, a HER2-directed antibody-drug conjugate, in patients with advanced HER2-mutant non-small-cell lung cancer (HORIZON-Lung): phase 2 results from a multicentre, single-arm study. Lancet Oncol 2025;26:437-46. [Crossref] [PubMed]
51. Xiong A, Wang L, Chen J, et al. Ivonescimab versus pembrolizumab for PD-L1-positive non-small cell lung cancer (HARMONi-2): a randomised, double-blind, phase 3 study in China. Lancet 2025;405:839-49. [Crossref] [PubMed]
52. Chen Z, Yang F, Jiang Z, et al. Ivonescimab plus chemotherapy versus tislelizumab plus chemotherapy as first-line treatment for advanced squamous non-small-cell lung cancer (HARMONi-6): a randomised, double-blind, phase 3 trial. Lancet 2025;406:2078-88. [Crossref] [PubMed]
53. Yang JC, Lu S, Hayashi H, et al. Overall Survival with Amivantamab-Lazertinib in EGFR-Mutated Advanced NSCLC. N Engl J Med 2025;393:1681-93. [Crossref] [PubMed]
54. Jänne PA, Planchard D, Kobayashi K, et al. Survival with Osimertinib plus Chemotherapy in EGFR-Mutated Advanced NSCLC. N Engl J Med 2026;394:27-38. [Crossref] [PubMed]
55. Yang JC, Wang M, Doucet L, et al. Phase II Dose-Randomized Study of Sunvozertinib in Platinum-Pretreated Non-Small Cell Lung Cancer With Epidermal Growth Factor Receptor Exon 20 Insertion Mutations (WU-KONG1B). J Clin Oncol 2025;43:3198-208. [Crossref] [PubMed]
56. Piotrowska Z, Passaro A, Nguyen D, et al. Zipalertinib in Patients With Epidermal Growth Factor Receptor Exon 20 Insertion-Positive Non-Small Cell Lung Cancer Previously Treated With Platinum-Based Chemotherapy With or Without Amivantamab. J Clin Oncol 2025;43:2387-97. [Crossref] [PubMed]
57. Le X, Kim TM, Loong HH, et al. Sevabertinib in Advanced HER2-Mutant Non-Small-Cell Lung Cancer. N Engl J Med 2025;393:1819-32. [Crossref] [PubMed]
58. Heymach JV, Ruiter G, Ahn MJ, et al. Zongertinib in Previously Treated HER2-Mutant Non-Small-Cell Lung Cancer. N Engl J Med 2025;392:2321-33. [Crossref] [PubMed]
59. Barlesi F, Yao W, Duruisseaux M, et al. Adagrasib versus docetaxel in KRAS(G12C)-mutated non-small-cell lung cancer (KRYSTAL-12): a randomised, open-label, phase 3 trial. Lancet 2025;406:615-26. [Crossref] [PubMed]
60. Sacher AG, Miller WH Jr, Patel MR, et al. Single-Agent Divarasib in Patients With KRAS G12C-Positive Non-Small Cell Lung Cancer: Long-Term Follow-Up of a Phase I Study. J Clin Oncol 2025;43:3249-53. [Crossref] [PubMed]
61. Zhong J, Zhao J, Duan J, et al. Glecirasib plus sitneprotafib in patients with KRAS(G12C)-mutated non-small-cell lung cancer in China: an open-label, multicentre, single-arm, phase 1/2a trial. Lancet Respir Med 2026;14:163-73. [Crossref] [PubMed]
62. Pérol M, Li W, Pennell NA, et al. Taletrectinib in ROS1+ Non-Small Cell Lung Cancer: TRUST. J Clin Oncol 2025;43:1920-9. [Crossref] [PubMed]
63. Johnson ML, Smit EF, Felip E, et al. Updated Overall Survival Analysis From the Phase II PHAROS Study of Encorafenib Plus Binimetinib in Patients With BRAF V600E-Mutant Metastatic Non-Small Cell Lung Cancer. J Clin Oncol 2025;43:3706-13. [Crossref] [PubMed]
64. Zhou C, Wang Z, Sun M, et al. Sugemalimab versus placebo, in combination with platinum-based chemotherapy, as first-line treatment of metastatic non-small-cell lung cancer (GEMSTONE-302): 4-year outcomes from a double-blind, randomised, phase 3 trial. Lancet Oncol 2025;26:887-97. [Crossref] [PubMed]
65. Sun H, Li M, Huang W, et al. Thoracic Radiotherapy Improves the Survival in Patients With EGFR-Mutated Oligo-Organ Metastatic Non-Small Cell Lung Cancer Treated With Epidermal Growth Factor Receptor-Tyrosine Kinase Inhibitors: A Multicenter, Randomized, Controlled, Phase III Trial. J Clin Oncol 2025;43:412-21. [Crossref] [PubMed]
66. Mountzios G, Sun L, Cho BC, et al. Tarlatamab in Small-Cell Lung Cancer after Platinum-Based Chemotherapy. N Engl J Med 2025;393:349-61. [Crossref] [PubMed]
67. Paulson KG, Lau SCM, Ahn MJ, et al. Safety and activity of tarlatamab in combination with a PD-L1 inhibitor as first-line maintenance therapy after chemo-immunotherapy in patients with extensive-stage small-cell lung cancer (DeLLphi-303): a multicentre, non-randomised, phase 1b study. Lancet Oncol 2025;26:1300-11. [Crossref] [PubMed]
68. Paz-Ares L, Borghaei H, Liu SV, et al. Efficacy and safety of first-line maintenance therapy with lurbinectedin plus atezolizumab in extensive-stage small-cell lung cancer (IMforte): a randomised, multicentre, open-label, phase 3 trial. Lancet 2025;405:2129-43. [Crossref] [PubMed]
69. Gondi V, Pugh SL, Mehta MP, et al. Hippocampal Avoidance During Prophylactic Cranial Irradiation for Patients With Small Cell Lung Cancer: Randomized Phase II/III Trial NRG-CC003. J Clin Oncol 2025;43:3516-25. [Crossref] [PubMed]
70. Aizer AA, Tanguturi SK, Shi DD, et al. Stereotactic Radiosurgery in Patients With Small Cell Lung Cancer and 1-10 Brain Metastases: A Multi-Institutional, Phase II, Prospective Clinical Trial. J Clin Oncol 2025;43:2986-97. [Crossref] [PubMed]
71. Hallet J, Rousseau M, Wakeam E, et al. Contemporary Incidence and Survival of Lung Neuroendocrine Neoplasms. JAMA Netw Open 2025;8:e2535125. [Crossref] [PubMed]
72. Tan J, Yang SC, Dinan MA, et al. Biomarker-Specific Survival and Medication Cost for Patients With Non-Small Cell Lung Cancer. JAMA Netw Open 2025;8:e2514519. [Crossref] [PubMed]
73. Pramesh CS, Kelly K, Hunter DJ, et al. The IASLC-Lancet Commission: a transformative global approach to lung cancer. Lancet 2025;406:1450-2. [Crossref] [PubMed]
74. Rivera MP, Benefield T, Durham DD, et al. Lung Cancer Screening in Cancer Survivors vs Those Without a History of Cancer. JAMA Netw Open 2025;8:e2535000. [Crossref] [PubMed]
75. Ma Z, Cao F, Liao M, et al. Cardiovascular adverse events associated with epidermal growth factor receptor tyrosine kinase inhibitors in EGFR-mutated non-small cell lung cancer: systematic review and network meta-analysis. BMJ 2025;390:e082834. [Crossref] [PubMed]