bims-meluca 2026-02-15 papers

on Metabolism of non-small cell lung carcinoma

Issue of 2026–02–15
ten papers selected by
the Muñoz-Pinedo/Nadal (PReTT) lab, L’Institut d’Investigació Biomèdica de Bellvitge

Genomic alterations and their correlation with metabolic-related genes in lung cancer.
EGFR and IRE1α pathways are associated with distinct immunomodulatory gene expression profiles in NSCLC cells with acquired resistance to EGFR TKIs.
Tumor-immune-neural circuit disrupts energy homeostasis in cancer cachexia.
Loss of PIK3CA allows in vitro growth but not in vivo progression of KRAS mutant lung adenocarcinoma in a syngeneic orthotopic implantation model.
B cell phenotypes and antibody signatures associate with interpatient variation in the lung adenocarcinoma tumor microenvironment.
Body Composition in Resectable Non-Small Cell Lung Cancer Patients With Preoperative Sarcopenia: A Propensity-Matched Analysis.
Differences in immunity and survival among non-small cell lung cancer patients: a gut microbiota perspective.
Influence of oncostatin M and leukemia inhibitor factor on sarcopenia in individuals with cancer and malnutrition.
MUC14 suppresses lung adenocarcinoma via integrin α8β6/PI3K/AKT/MAPK modulating cisplatin response and immunity.
Prognostic value of baseline 18F-FDG PET/CT metabolic parameters combined with clinical and pathological features in surgically resected ALK-positive non-small cell lung cancer.

Clin Transl Oncol. 2026 Feb 11.

Genomic alterations and their correlation with metabolic-related genes in lung cancer.

Gauri Gaur, Niraj Kumar Jha, Lokesh Gambhir, Seema V Acharya, Dhruv Kumar.

  Non-small cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality worldwide, with 5-year survival rates below 21% primarily due to therapeutic resistance and metastatic progression. Genomic alterations in KRAS, EGFR, TP53, and MYC drive metabolic reprogramming that sustains tumor proliferation and therapy resistance. This review synthesizes evidence linking specific genomic alterations, including variant-specific KRAS alleles (G12C, G12D, and G12V) and TP53 gain- or loss-of-function mutations, to distinct metabolic phenotypes in NSCLC. It further examines the immunometabolic consequences of co-occurring mutations such as KRAS with TP53 or STK11/LKB1. The literature synthesis integrates genomic, metabolic, and immunologic profiling data to identify mutation-specific metabolic vulnerabilities and therapeutic targets. Genomic alterations establish distinct metabolic dependencies: KRAS-driven tumors exhibit enhanced glycolysis and glutaminolysis, EGFR-mutant tumors demonstrate increased lipogenesis, and TP53 loss promotes metabolic flexibility. Accumulation of lactate and depletion of glucose in the tumor microenvironment suppress CD8+ T-cell function, facilitating immune evasion. Rational combination strategies that pair genomic-targeted agents (sotorasib and adagrasib) with metabolic inhibitors (CB-839 and TVB-2640) show promise in overcoming adaptive resistance. Integrating genomic and metabolic profiling may enhance precision oncology approaches and improve clinical outcomes.

Keywords:  Genomic alterations; Glycolysis; Immunometabolism; Lung cancer; Metabolic reprogramming; Tumor microenvironment

DOI:  https://doi.org/10.1007/s12094-026-04229-4
Arch Biochem Biophys. 2026 Feb 09. pii: S0003-9861(26)00033-0. [Epub ahead of print] 110762

EGFR and IRE1α pathways are associated with distinct immunomodulatory gene expression profiles in NSCLC cells with acquired resistance to EGFR TKIs.

Yi-Shiuan Wang, Hsiu-Chuan Chou, En-Chi Liao, Hsin-Yi Chen, Meng-Wei Lin, Yu-Shan Wei, Li-Hsun Lin, Yi-Ru Chang, Han Meng, Yueh-Feng Wen, Hong-Lin Chan.

  EGFR-TKI-resistant cell lines were established by long-term exposure to gefitinib, afatinib, and osimertinib via the PC9 model. This model helps study EGFR-TKI resistance mechanisms in non-small cell lung cancer (NSCLC) and may offer insights to improve treatment outcomes, particularly for patients unresponsive to anti-PD-1/PD-L1 therapies. We investigated molecular alterations in these resistant cell lines using multi-omics techniques, including genome sequencing, proteomics, and transcriptomics. Differential dependencies on EGFR downstream pathways were observed among resistant cell lines. Immune evasion mechanisms were analyzed to identify alterations in EGFR downstream pathways and unfolded protein response (UPR) elements contributing to immune-related transcriptional patterns. Afatinib-resistant cells exhibited the most pronounced immunosuppressive-like transcriptional features compared to gefitinib- and osimertinib-resistant cells. This was characterized by elevated PD-L1 expression, minimal changes in MHC class 1 levels, down-regulation of shared immune-modulated gene sets, and a cytokine profile favoring immunosuppression (e.g., lower IL-6, CXCL10, IFN-γ, CCL22; higher IL-8). These findings suggest a poor prognosis and limited efficacy of anti-PD-1/PD-L1 immunotherapy in NSCLC patients with similar resistance profiles. Notably, inhibition of IRE1α signaling partially reversed this immunosuppressive-like transcriptional signature, indicating a potential strategy to restore immune responsiveness in afatinib-resistant NSCLC. Our study underscores the distinct remodeling effects of different EGFR-TKIs on tumor-cell-intrinsic immunoregulatory pathways. Targeting endoplasmic reticulum (ER) stress pathways alongside immune checkpoint inhibitors may be crucial for overcoming resistance mechanisms identified here. These insights provide a rationale for personalized treatment strategies tailored to the immune-related gene-expression profiles observed in EGFR-TKI-resistant NSCLC models, aiming to enhance therapeutic responses and improve clinical outcomes.

Keywords:  2D-DIGE; EGFR-TKI; Immunomodulatory; RNA-sequencing

DOI:  https://doi.org/10.1016/j.abb.2026.110762
Cancer Cell. 2026 Feb 12. pii: S1535-6108(26)00053-X. [Epub ahead of print]

Tumor-immune-neural circuit disrupts energy homeostasis in cancer cachexia.

Xiuhui Shi, Alex X Arreola, Zhijun Zhou, Jingxuan Yang, Mingyang Liu, Yang Cai, Yu Ren, Hao Yuan, Qun Chen, Xinjie Chen, Xinyu Yang, Yimei Meng, Jingyi Wang, Wenyi Luo, Michael C Rudolph, Rohan Varshney, Kar-Ming Fung, Chao Xu, Wei R Chen, Michael S Bronze, Lei Zheng, Yi-Ping Li, Courtney W Houchen, Yuqing Zhang, Min Li.

  Cancer-induced cachexia and anorexia are debilitating complications across many cancers, yet effective treatments remain limited due to a poor understanding of the underlying mechanisms. Here, we identify an uncharacterized tumor-immune-neural circuit driving these syndromes, centered on growth and differentiation factor 15 (GDF15). Using genetically engineered mouse models, we find that loss of GDF15 protects against appetite loss, muscle wasting, and fat loss in pancreatic, lung, and skin cancers. Single-cell RNA sequencing reveals macrophages as a major source of GDF15, induced by tumor-derived colony-stimulating factor 1 (CSF1). GDF15 acts via the central nervous system to enhance β-adrenergic signaling in the tumor microenvironment, thereby amplifying cachexia. The disruption of this feedforward loop with GDF15-neutralizing antibody, anti-CSF1R antibody, or Rearranged during Transfection (RET) inhibitor markedly reduces both cachexia and anorexia. These findings reveal a non-cell-autonomous mechanism linking tumor signals, macrophage-derived GDF15, and neural pathways, highlighting the tumor-immune-neural triad as a promising therapeutic target.

Keywords:  adipose loss; body composition; energy expenditure; hormone; metabolic stress; muscle atrophy; norepinephrine; sympathetic nerve; tumor immune microenvironment; tumor-associated macrophages

DOI:  https://doi.org/10.1016/j.ccell.2026.01.014
bioRxiv. 2026 Feb 04. pii: 2026.02.02.701385. [Epub ahead of print]

Loss of PIK3CA allows in vitro growth but not in vivo progression of KRAS mutant lung adenocarcinoma in a syngeneic orthotopic implantation model.

Abigail L Booth, Giuseppe Caso, Barbara Rosati, Ya-Ping Jiang, Wei-Xing Zong, Richard Z Lin, Harold Bien.

Constitutively active KRAS mutations are highly prevalent in lung cancers, but the direct role of its downstream phosphatidylinositol 3-kinase (PI3K) pathway in tumor progression remains unclear. A previous study established the requirement for PIK3CA, the alpha catalytic isoform, in lung tumor development in mouse models with an intact Trp53 tumor suppressor. In this study, we further investigated the requirement for PIK3CA for tumor growth both in vitro and in vivo . We first generated a "KPA" cell line by genetically deleting Pik3ca from a murine lung adenocarcinoma "KP" cell line harboring oncogenic Kras G12D and lacking Trp53 . We found that Pik3ca is not required for cell survival and growth in vitro , even under anchorage-independent conditions but reduced the growth rate by 20%. We next orthotopically implanted KP and KPA cells into syngeneic mice and found that PIK3CA is absolutely required for tumor progression, even in the absence of Trp53 . Implantation of KP cells, or a "KPS" cell line lacking the Stk11 gene, led to rapid tumor growth and death of all host animals. In contrast, mice implanted with KPA cells all survived with no detectable lung tumors. The gene expression profiles from cultured cell lines suggest KPA cells may be vulnerable to oxidative stress. Indeed, we found KPA cells were more sensitive to hydrogen peroxide and diethyl maleate-induced oxidative stress as compared to KP and KPS cells. Together, these results demonstrate that PIK3CA is not required for lung cancer cell growth induced by mutant KRAS in vitro but is critically needed for in vivo progression and growth.

DOI: https://doi.org/10.64898/2026.02.02.701385
Front Immunol. 2025 ;16 1739637

B cell phenotypes and antibody signatures associate with interpatient variation in the lung adenocarcinoma tumor microenvironment.

Kamille M Rasche, David Tieri, Samantha M Morrissey, Hong Li, Fan Zhang, William S Gibson, Easton E Ford, Uddalok Jana, Melissa L Smith, Jun Yan, Corey T Watson.

   Introduction: Non-small cell lung cancer is the leading cause of cancer-related mortality worldwide, with lung adenocarcinoma (LUAD) as the most common subtype. Although early-stage disease is often treated surgically, advanced LUAD typically requires chemotherapy, radiation, and/or immunotherapy, largely focused on T cell-mediated responses. Therapeutic efficacy, however, is also shaped by tumor-infiltrating (TI) B cells, whose roles in LUAD remain incompletely understood.
Methods: We performed cytometry by time of flight (CyTOF) using 44 markers on matched tumor, adjacent lung, and peripheral blood samples from 48 LUAD patients to define TI immune landscapes. 66 immune cell subsets were identified, and patients were stratified into four groups based on TI cell composition. Adaptive immune receptor repertoire sequencing (AIRR-seq) of IgM and IgG was conducted on matched samples from 29 patients to assess clonal expansion and affinity maturation. Subisotype-resolved AIRR-seq was additionally performed on tumor samples from 18 patients.
Results: CyTOF analysis revealed four patient groups with distinct TI immune profiles. AIRR-seq demonstrated increased clonal expansion and affinity maturation in tumors compared to adjacent lung and blood. Tumor-specific IGHV enrichment patterns were observed but were not associated with patient group assignment. Instead, clonal expansion was greatest in tumors with higher lymphocyte proportions. Subisotype-resolved analysis showed enrichment of IGHG2 and IGHG3 in tumors from patients with low TI B cell abundance, whereas IGHG4 was enriched in patients with high TI B cell infiltration and correlated with four CyTOF-defined immune subsets.
Discussion: These findings reveal substantial inter-individual variation in TI immune landscapes and highlight distinct B cell repertoire and subisotype features within LUAD tumors. Together, these data suggest that B cell composition and antibody subisotype usage may contribute to immune contexture and could inform the development of more tailored immunotherapeutic strategies in LUAD.

Keywords:  B cells; antibody repertoire; cytometry by time-of-flight; lung adenocarcinoma; sequencing

DOI:  https://doi.org/10.3389/fimmu.2025.1739637
J Cachexia Sarcopenia Muscle. 2026 Feb;17(1): e70230

Body Composition in Resectable Non-Small Cell Lung Cancer Patients With Preoperative Sarcopenia: A Propensity-Matched Analysis.

Zhu-Zhu Wang, Jin-Yu He, Rui-Feng Tang, Tian Zhang, Le Luo, Yu-Han Zhang, Jing Zhang, Xiao-Qing Lv, Jing-Fang Hong.

   BACKGROUND: Sarcopenia is a prognostic factor in patients with early-stage non-small cell lung cancer (NSCLC). Previous studies have focused on muscle loss, yet comprehensive body composition (BC) alterations in this population remain poorly characterized. This study aimed to delineate BC alterations using bioelectrical impedance analysis (BIA) in early-stage NSCLC patients with preoperative sarcopenia, with a specific focus on sex-specific disparities and the identification of independent BC factors associated with sarcopenia.
METHODS: In this case-control study, 460 patients with Stage I-II NSCLC were initially enrolled from two tertiary hospitals in Anhui, China. Sarcopenia was diagnosed preoperatively based on Asian Working Group for Sarcopenia 2019 criteria. Multifrequency BIA was performed within 48 h before surgery. Propensity score matching (PSM) at a 1:4 ratio was applied to balance covariates (age, sex, height, physical activity, nutritional status, clinical stage, histology, extent of resection and diabetes). Multivariable logistic regressions were used to examine the associations between BC parameters and sarcopenia.
RESULTS: After PSM, 47 sarcopenic and 162 nonsarcopenic patients were well-matched. Sarcopenic patients exhibited systemic depletion beyond reduced muscle mass, including lower body fat mass (BFM: 12.70 vs. 18.60 kg), body cell mass (BCM: 23.90 vs. 29.10 kg), bone mineral content (BMC: 2.22 vs. 2.52 kg = 0.002), basal metabolic rate (BMR: 1151.00 vs. 1340.50 kcal) and elevated extracellular water/intracellular water ratio (ECW/ICW: 0.64 vs. 0.63; all p < 0.001). Distinct sex-specific phenotypes were identified: Sarcopenic females demonstrated coordinated reductions in muscle, fat and minerals (appendicular skeletal muscle mass [ASM]: 13.10 vs. 15.43 kg; BFM: 11.20 vs. 19.20 kg; percent body fat [PBF]: 25.41% vs. 32.88%; BMC: 2.05 vs. 2.26 kg; all p < 0.001), whereas males exhibited isolated muscle loss with preserved adiposity and minerals (PBF: 23.04% vs. 24.29%, p = 0.463; BMC: 2.64 vs. 2.80 kg, p = 0.141). In the fully adjusted model, ASM (OR = 0.03, 95%CI: 0.01-0.08), BFM (OR = 0.81, 95%CI: 0.75-0.87) and BMR (OR = 0.95, 95%CI: 0.94-0.97) were independent factors associated with sarcopenia. Sex significantly modified associations for waist-hip ratio, soft lean mass, FFM and PBF (all p for interaction < 0.05). Sensitivity analyses based on sarcopenia severity supported the robustness of the primary findings.
CONCLUSIONS: Preoperative sarcopenia in early-stage NSCLC involves multicompartment depletion accompanied by cellular dysfunction and metabolic impairment, exhibiting distinct sex-specific phenotypes. BIA provides a practical tool for multidimensional BC assessment, necessitating to integrate it into clinical workflows and developing sex-tailored prehabilitation strategies.

Keywords:  bioelectrical impedance analysis; body composition; non‐small cell lung cancer; propensity score matching; sarcopenia

DOI:  https://doi.org/10.1002/jcsm.70230
Am J Transl Res. 2026 ;18(1): 444-454

Differences in immunity and survival among non-small cell lung cancer patients: a gut microbiota perspective.

Hao Liao, Xiangping Luo, Liqin Jiang.

   OBJECTIVE: This study aimed to investigate the relationship between key gut microbiota [Enterococcus, Escherichia coli (E. coli), Bifidobacterium, and Lactobacillus] and immune function in Chinese patients with non-small cell lung cancer (NSCLC).
METHODS: This study included 208 patients with NSCLC enrolled between March 2021 and June 2023. Fecal samples were collected from patients for quantitative analysis of Enterococcus, E. coli, Bifidobacterium, and Lactobacillus. Additionally, peripheral blood samples were collected from patients, and levels of T lymphocyte subsets (CD3+, CD4+, and CD8+) were measured using flow cytometry. Analysis was conducted based on 6-month immune checkpoint inhibitor (ICI) efficacy and survival outcomes to examine the relationship between gut microbiota, immune function, and prognosis in NSCLC patients. Pearson correlation coefficients were used to investigate the interrelationships between microbial abundance and immune variables.
RESULTS: In the immune checkpoint inhibitor (ICI) responders (R group), higher proportions of Bifidobacteria and Lactobacilli were observed, whereas non-responders (NR group) exhibited increased proportions of Enterococcus and E. coli. Furthermore, Bifidobacteria and Lactobacilli showed positive correlations with T cell counts but negative correlations with inflammatory cytokine levels. Opposing relationships were seen for Enterococcus and E. coli, which correlated negatively with T cells and positively with IL-6 and TNF-α (P < 0.05).
CONCLUSION: In NSCLC, Bifidobacterium and Lactobacillus promote beneficial immune feedback loops by activating T cells and exerting anti-inflammatory effects, thereby supporting antitumor immunity.

Keywords:  NSCLC; gut microbiota; immune cells; inflammatory factors; prognosis

DOI:  https://doi.org/10.62347/OIHD4474
Endocrinol Diabetes Nutr (Engl Ed). 2026 Feb;pii: S2530-0180(25)00165-9. [Epub ahead of print]73(2): 501677

Influence of oncostatin M and leukemia inhibitor factor on sarcopenia in individuals with cancer and malnutrition.

Daniel de Luis, David Primo, Olatz Izaola, Juan José López Gómez.

   OBJECTIVES: Oncostatin-M (OSM) and leukemia inhibitor factor (LIF) have shown relationship with muscle catabolism in Disease Related Malnutrition (DRM). We aimed to investigate the role of OSM and LIF in sarcopenia among patients with cancer and DRM.
DESIGN AND METHODS: 33 cancer patients with DRM were included. Muscle mass assessment via ultrasound at the rectus femoris (RF) level, bioimpedance (skeletal muscle mass (SMM), appendicular SMM (aSMM), and the aSMM index (aSMMI)), and circulating OSM and LIF levels were evaluated. Confirmed sarcopenia was diagnosed based on the criteria of low muscle strength combined with an abnormal aSMMI.
RESULTS: The average age was 66.5±15.5 years. The gender distribution was 16 females (48.5%) and 17 males (51.5%). Sixteen patients (48.5%) were diagnosed with sarcopenia. Patients with sarcopenia showed worse values: phase angle (-0.4±0.1°; p=0.01), reactance (-4.9±1.8Ω; p=0.03), SMM (-2.0±0.3kg; p=0.03), aSMM (-1.6±0.2kg; p=0.03), aSMMI (-0.6±0.2kg/m2; p=0.02), muscle area RF (-0.7±0.2cm2; p=0.04), Y-axis RF (-0.4±0.1cm; p=0.03), muscle strength (-8.1±1.5kg; p=0.01) and albumin levels (-0.3±0.1g/dl; p=0.03). Circulating levels of OSM and LIF were elevated in patients with sarcopenia, with values of 10.6±1.3pg/ml (p=0.01) and 6.4±2.3pg/ml (p=0.02), respectively. Logistic regression analysis indicated a significant OR (odds ratio) of sarcopenia associated with elevated median levels of OSM (OR=3.6, 95% CI=1.4-9.9; p=0.01) and LIF-1 (OR=3.3, 95% CI=1.3-10.1; p=0.01) CONCLUSION: High levels of serum OSM and LIF-1 were closely associated with sarcopenia in patients with cancer and DRM.

Keywords:  Cancer; Cáncer; LIF; Malnutrición; Malnutrition; Oncostatin M; Oncostatina M

DOI:  https://doi.org/10.1016/j.endien.2025.501677
Sci Rep. 2026 Feb 08.

MUC14 suppresses lung adenocarcinoma via integrin α8β6/PI3K/AKT/MAPK modulating cisplatin response and immunity.

Xiaoqing Li, Ming Li, Shizhuan Huang, Zhihua Zhang, Chen Xing, Shan Yu, Guiping Han.

  MUC14/Endomucin, a transmembrane mucin, is a potential prognostic biomarker in malignancies. This study aimed to elucidate the functional impact of MUC14 on tumor proliferation, migration, immune microenvironment modulation, and cisplatin response in lung adenocarcinoma (LUAD), and investigate its molecular mechanisms. LUAD cell lines with MUC14 overexpression (MUC14-OE) or silencing were constructed. Malignant behaviors were assessed via CCK-8, Transwell, and colony formation assays. Immune cell infiltration was quantified by CD3+/CD8 + immunohistochemistry. Subcutaneous xenograft and tail-vein metastasis murine models evaluated in vivo tumor progression and cisplatin responsiveness. Mechanisms were characterized using FRET and western blotting. Multiplatform bioinformatics analysis of public databases correlated MUC14 expression with clinical outcomes, immune infiltration, and chemotherapy response. MUC14-OE inhibited LUAD cell proliferation, migration, colony formation, and adhesion, while silencing promoted these phenotypes. MUC14 expression positively correlated with CD3+/CD8 + T-cell infiltration. In vivo, MUC14-OE suppressed subcutaneous tumor growth, lung metastasis, and enhanced cisplatin efficacy. Mechanistically, MUC14 inhibited integrin α8β6 clustering, suppressing PI3K/AKT and MAPK/ERK signaling. Cisplatin sensitization involved JNK/c-Jun pathway activation. This study establishes MUC14 as a multifunctional tumor suppressor in LUAD. It inhibits integrin α8β6-mediated PI3K/AKT and MAPK/ERK signaling to suppress tumor growth, promotes CD8+ T-cell infiltration, and augments cisplatin sensitivity via the JNK/c-Jun pathway. These findings nominate MUC14 as a prognostic biomarker and therapeutic target, suggesting combinatorial strategies integrating immunotherapy and chemotherapy.

Keywords:  Cisplatin; Immunotherapy; Integrin; Lung adenocarcinoma; MUC14

DOI:  https://doi.org/10.1038/s41598-026-39019-z
Lung Cancer. 2026 Jan 29. pii: S0169-5002(26)00042-5. [Epub ahead of print]213 108948

Prognostic value of baseline 18F-FDG PET/CT metabolic parameters combined with clinical and pathological features in surgically resected ALK-positive non-small cell lung cancer.

You Cheng, Jian-Jiang Huang, Hao-Yu Zhu, Dan Shao, Hu-Bing Wu.

   OBJECTIVE: To investigate the prognostic value of baseline 18F-fluorodeoxyglucose (18F-FDG) PET metabolic parameters, along with clinical and pathological characteristics, in predicting postoperative outcomes in patients with ALK-positive non-small cell lung cancer (NSCLC).
METHODS: A retrospective analysis was conducted on patients at our institution with pathologically confirmed ALK-positive NSCLC. Baseline PET metabolic parameters, clinical characteristics, and pathological features were examined. Receiver operating characteristic (ROC) curve analysis was performed to determine the optimal cutoff values for all parameters. Survival analyses,including Kaplan-Meier curves, the log-rank test, and Cox proportional hazards regression, were employed to assess disease-free survival (DFS) and identify independent prognostic indicators.
RESULTS: The analysis included 78 participants with a median follow-up time of 38.5 months (95% CI: 28.4 - 48.6). The median DFS was 72.8 months (95% CI: 44.7 - 100.8). Univariate analysis revealed significant associations between DFS and several clinical (T stage, overall clinical stage, and CYFRA21-1), PET (SUVmax, SUVmean, SUVpeak, TLG, and MTV), and pathological (Ki-67 index, tumor spread through air spaces [STAS], and pleural invasion) factors (p < 0.05, for all). Multivariate Cox regression analysis identified the following independent predictors of DFS: SUVmax (HR = 16.152, p = 0.002), STAS (HR = 6.122, p = 0.040), T stage (HR = 2.588, p = 0.049), and preoperative CYFRA21-1(HR = 6.509, p = 0.028).
CONCLUSION: The assessment of 18F-FDG PET metabolic parameters, pathological factors, and clinical characteristics provides independent prognostic information for postoperative outcomes in patients with ALK-positive NSCLC. These findings may help inform postoperative adjuvant treatment strategies.

Keywords:  ALK-positive; Non-small cell lung cancer(NSCLC); PET metabolic parameters; Pathological parameters

DOI:  https://doi.org/10.1016/j.lungcan.2026.108948