Background: Low-dose computed tomography (LDCT) screening has been proven to reduce mortality in heavy smokers, there are limited evidence for its effectiveness in individuals with non-smoking risk factors, who represent a significant proportion of lung cancer patients in East Asia. We evaluated 10-year mortality and stage-shift outcomes in the China Lung Cancer Screening (CLUS) trial. Methods: From November 2013 to November 2014, eligible participants with high-risk factors of lung cancer were randomly assigned to undergo three biennial screenings or a control group with questionnaire inquiries. Data were collected on cases of lung cancer and deaths by the Shanghai Municipal Center for Disease Control and Prevention. Median follow-up was estimated using the reverse Kaplan–Meier method, and lung cancer mortality was compared using Fine–Gray competing-risk models. Results: The median follow-up time was 11.7 years. LDCT screening increased the detection of early-stage lung cancers, and only two interval cancers occurred among those on biennial screening. After adjusting for baseline age and sex, the screening group showed a 56% reduction in the risk of lung-cancer death compared with the control group (Adjusted sHR = 0.44; 95% CI, 0.21 to 0.90; p = 0.026). The mortality benefit was consistent across smoking-status subgroups (p for interaction = 0.760), despite non-significance in never-smokers. In exploratory analyses, participants with high exposure to cooking oil fume and those with a family history of malignancy appeared to derive a greater mortality benefit from LDCT screening. Conclusion: Biennial LDCT screening significantly reduced 10-year lung cancer mortality and promoted early-stage diagnosis in a Chinese population selected using multiple risk factors. The findings support a screening framework extending beyond tobacco exposure and identify potentially informative risk factors. This community-based biennial strategy may improve feasibility, coverage, and resource efficiency and provides long-term evidence for developing individualized lung cancer screening programs in China.

Table S1. Cell-type annotations for the scRNA data

Insufficient nutrient and oxygen supply remains a critical barrier to cell survival and tissue regeneration in threedimensional scaffolds. Here, we developed a self-supplying glucose/oxygen hydrogel (SGO-Gel) that intrinsically provides sustained metabolic support for dental pulp stem cells (DPSCs). The hydrogel was constructed by crosslinking oxidized dextran with phenylboronic acid-grafted carboxymethyl chitosan through dynamic Schiff-base and phenylboronic ester bonds, yielding an injectable, self-healing network capable of releasing glucose upon dextranase degradation. Polydopamine-coated calcium peroxide nanoparticles were incorporated to achieve controlled oxygen release while maintaining structural stability. SGO-Gel exhibited a uniform porous architecture that facilitated nutrient diffusion, as confirmed by molecular dynamics simulations. In vitro, the hydrogel significantly enhanced DPSCs survival under combined glucose starvation and hypoxia, maintaining viability above 70 % after 7 days compared to less than 30 % in control gels. Transcriptomic analysis revealed activation of the MAPK/Erk1/2 pathway and upregulation of VEGF, which was consistent with functional assays showing enhanced endothelial migration and tube formation. In an ectopic transplantation model, SGO-Gel promoted the formation of vascularized pulp-like tissue with abundant collagen deposition and odontoblast differentiation. Overall, SGO-Gel shifts the paradigm from passive scaffolds to metabolically active matrices, offering a promising strategy for dental pulp regeneration and broader tissue engineering applications where ischemia and limited vascularization remain key challenges.

Table S2. scRNA cell-type markers.

Objective Using ovarian cancer datasets from public databases,identify copper death-related genes in ovarian cancer tissues and construct a clinical prognostic risk scoring model for ovarian cancer patients based on these genes.Methods We downloaded the OC data of TCGA,the GSE26193,GSE63885 dataset from GEO and retrieved 10 cuproptosis related genes(CRGs)and analyzed their chromosomal localization,expression correlation,and mutation patterns based on the datasets.Using these genes,we clustered the OC samples to identify different molecular subtypes of copper induced death.We analyzed the differential genes and functional enrichment between different subtypes and obtained feature genes with predictive ability for prognosis through survival regression analyses.Based on these feature genes,we constructed a risk scoring model and incorporated the clinical characteristics ofpatients to jointly predict their survival rate.Results In ovarian cancer samples,10 copper death-related genes can stably divide the samples into two molecular subtypes,and there are significant differences in clinical and immune characteristics and drug sensitivity between them.After further screening,seven prognostic genes(RARRES1、CXCL10、PI3、CXCL11、THEMIS2、GBP2、RPL39L)were obtained,and the risk model based on them combined with age predicted that the AUC of patients'1-,3-,and 5-year survival rates were all greater than 0.7,showing good clinical application prospects.Conclusion The mechanism of cuproptosis and its key genes might become therapeutic targets for ovarian cancer.The subtypes of cuproptosis provide a theoretical basis for personalized clinical treatment.The predictive model constructed by key prognostic genes has promising clinical application effects.

Table S3. Clinical phenotypes of 167 patients.

Table S6. The complete list of somatic mutations.

The fat mass and obesity-associated protein (FTO), a pivotal eraser of N6-methyladenosine (m6A) modifications, has emerged as a critical biomarker and therapeutic target in tumorigenesis. However, the accurate assessment of FTO activity in biological settings remains challenging due to limitations in sensitivity and cellular compatibility of existing methods. Herein, we develop an enzyme-free DNA circuit that integrates m6A-directed deoxyribozyme (DNAzyme) activation with autocatalytic hybridization assembly (AHA) for the sensitive monitoring of demethylase activity. In this system, FTO-mediated demethylation triggers the DNAzyme, which via a docking hairpin, transduces the recognition event into a cross-catalytic hybridization chain reaction (HCR)-catalytic hairpin assembly (CHA) circuit, thereby generating exponential signal amplification. This biosensing platform achieves an exceptional detection limit of 22.6 pM and is successfully applied to the screening of FTO inhibitors. Furthermore, the platform demonstrates the capability to reliably discriminate between cancer cells and normal cells based on distinct fluorescence signals. As a robust and entirely nucleic acid-based programmable strategy, this method provides a versatile and powerful tool for advancing epigenetic research and point-of-care molecular diagnostics.

Background: Highly sensitive nucleic acid detection is essential for analytical applications. Conventional methods often require complex pre-amplification procedures, limiting their practical utility in screening. Developing nucleic acid detection strategies with high sensitivity and selectivity, without target gene pre-amplification, remains a significant challenge. Results: This study integrates CRISPR/Cas12a recognition, catalytic hairpin assembly (CHA) amplification, and surface-enhanced Raman spectroscopy (SERS). CRISPR/Cas12a recognizes target nucleic acids and cleaves single-stranded DNA (ssDNA), thereby blocking the toehold-mediated strand displacement reaction (TSDR) and triggering CHA. Hairpin probe HP1 with C-Ag+-C structures bridges CHA, releasing Ag+ through cyclic amplification. Ag+ induces charge transfer and aggregation of AgNPs@4-ABT, generating strong SERS signals. The platform achieved femtomolar sensitivity and high selectivity in detecting pCaMV35S, with 96.4% accuracy in maize seeds and 100% in maize leaves. Significance: This strategy eliminates the need for pre-amplification of target genes by combining CRISPR's targeting feature, CHA, and ultrasensitive SERS detection. It demonstrates excellent performance in genetically modified organism screening, seed quality testing, and leaf sample analysis, providing a promising tool for food safety and agricultural regulation.

Table S7. The complete list of copy number alterations.

Cancer-associated fibroblasts (CAFs) are crucial for tumor microenvironment remodeling and tumor metastasis. Eukaryotic elongation factor 1 gamma (EEF1G) is aberrantly expressed in many tumors, but its role in lung cancer metastasis, particularly via exosome-mediated CAF activation, remains unclear. This study investigated EEF1G’s mechanism in lung cancer metastasis through exosome-induced tumor–stroma interactions. We performed bioinformatics analyses, single-cell RNA sequencing, and in vitro experiments (cell culture, exosome isolation, Western blot, qRT-PCR, and immunofluorescence) to characterize the function of EEF1G. A tumor-fibroblast co-culture system and in vivo metastasis models validated the EEF1G–exosome–CAF axis and associated signaling. EEF1G was overexpressed in lung cancer tissues and exosomes, which correlated with poor prognosis and CAF infiltration. EEF1G promoted the proliferation, migration, and invasion of lung cancer cells. Mechanistically, tumor cells transferred exosomal EEF1G to fibroblasts, inducing CAF transformation via the activation of the NF-κB/ROS pathway, increased mitochondrial activity, and IL-6 secretion. Subsequently, CAF-derived IL-6-rich exosomes activated STAT3 in tumor cells, enhancing invasiveness. In vivo experiments confirmed that this exosome-mediated bidirectional communication robustly facilitates lung cancer metastasis. This study elucidates a novel EEF1G exosome-mediated pathway that activates CAFs, establishing a pro-oncogenic feedback network. We emphasize the critical NF-κB/ROS/IL-6/STAT3 signaling axis in tumor–stroma interactions, offering potential therapeutic targets for lung cancer.

Table S5. Mutational profiles of 159 patients.

Lysosomal membrane permeabilization (LMP) mediated by pH-responsive materials has demonstrated considerable potential in tumor therapy. However, tumor cells exhibit a pronounced adaptive capacity to remodel lysosomal pH, thereby resisting LMP induction by pH-responsive materials and ultimately compromising therapeutic efficacy. To overcome this limitation, we engineered proton-driven π-π stacked copper-phycocyanin nanoparticles (CuPC NPs) designed to selectively promote LMP in tumor cells through their aggregation within acidic lysosomes. Mechanistically, aggregated CuPC NPs directly engage the RIPK3/p-MLKL signaling axis to trigger LMP, thereby orchestrating the synergistic activation of necroptosis, ferroptosis, and cuproptosis. Due to the coordinated induction of multimodal cell death, these pH-responsive CuPC NPs effectively inhibit primary breast tumor growth and suppress pulmonary metastasis. Collectively, our study establishes a strategy for precise modulation of lysosomal function to enhance antitumor efficacy and provides valuable insights into the development of lysosome-targeted nanotherapeutics for tumor treatment.

Superoxide anion (O-2(center dot-)) is a key reactive oxygen species involved in oxidative stress-related diseases and is typically overproduced in tumor cells, making it an important biomarker for cancer diagnosis. Herein, we report a near-infrared (NIR) turn-on fluorescent molecular probe, Cy- O-2(center dot-), for selective imaging of O-2(center dot-) and precise tumor identification. Cy-O-2(center dot-) is rationally designed based on a hemicyanine fluorophore, in which a trifluoromethanesulfonate group serves as a specific O-2(center dot-)-responsive recognition unit. In its intact form, fluorescence is effectively suppressed by blocking the intramolecular charge transfer (ICT) process. Upon reaction with O-2(center dot-), nucleophilic attack at the sulfonyl sulfur induces O-S bond cleavage, releasing the free hemicyanine fluorophore, thereby recovering the NIR fluorescence signal (similar to 742 nm). Cy-O-2(center dot-) exhibits high sensitivity, excellent selectivity toward O-2(center dot-) over other reactive species, and good biocompatibility. Cellular experiments demonstrate its capability to monitor endogenous and exogenous O-2(center dot-) fluctuations. More importantly, in vivo imaging in tumor-bearing mouse models confirms precise tumor visualization, highlighting its potential for early tumor diagnosis.

Table S4. Molecular subtypes of 406 RNA-seq samples.

Lung squamous cell carcinoma (LUSC) is a subtype of non-small cell lung cancer (NSCLC). Compared to lung adenocarcinoma (LUAD), LUSC is characterized by a greater propensity for recurrence and metastasis, poorer prognosis, and shorter survival. Therefore, further research into the pathogenesis of LUSC and the identification of new therapeutic targets are essential to advance clinical treatment options for this aggressive cancer. FAM111B, a serine protease and cancer-associated nuclear protein, has been implicated in various cancers. Previous studies have shown that FAM111B is closely associated with the progression of LUAD. However, our pan-cancer analysis suggests that FAM111B is highly expressed in LUSC and plays an oncogenic role, which is consistent with its higher expression in LUSC compared to LUAD in clinical samples. To investigate the functional role of FAM111B in LUSC, we developed both in vitro and in vivo knockdown models. Our results demonstrate that knocking down FAM111B significantly inhibits the proliferation, migration, and invasion of LUSC cells. Additionally, FAM111B knockdown induces cell cycle arrest in the S phase, further underscoring its role in LUSC progression. Mechanistically, FAM111B promotes LUSC migration and invasion by facilitating epithelial-mesenchymal transition (EMT). Moreover, the proliferation and cell cycle processes in LUSC may be regulated through the PI3K signaling pathway. In conclusion, our study elucidates the clinical relevance and molecular mechanisms of FAM111B in LUSC, highlighting its potential as a novel therapeutic target for this challenging cancer subtype.

Although first-line immunotherapy plus chemotherapy has made substantial progress in extensive-stage small-cell lung cancer (ES-SCLC), the survival benefit remains limited. DURABLE was a prospective, multicenter, open-label, randomized, phase II trial (NCT04985851) that evaluated the efficacy and safety of durvalumab plus anlotinib as consolidative maintenance therapy following first-line durvalumab plus platinum-etoposide chemotherapy in ES-SCLC. The primary endpoint was blinded independent central review-assessed progression-free survival (PFS) from randomization, with a two-sided p-value of <0.20 considered statistically significant. 66 patients were randomly assigned to maintenance therapy with durvalumab plus anlotinib (n = 34) or durvalumab alone (n = 32). Durvalumab plus anlotinib significantly improved PFS compared with durvalumab alone, with median PFS from randomization of 5.4 months versus 1.9 months (HR = 0.64; 80% CI, 0.44-0.94; p = 0.12). The incidence of grade 3-4 treatment-related adverse events in the combination and monotherapy groups was 24.2% and 12.5%, respectively. Patients with impaired antigen presenting capacity or low bTMB tended to show improved outcomes with combined maintenance therapy. While similar efficacy between the two groups was observed in patients with high antigen presenting capacity or high bTMB. These findings suggest that durvalumab plus anlotinib might be an effective and well-tolerated maintenance treatment option in ES-SCLC.

Identification of ischemic stroke during hyperacute phase is crucial for subsequent treatment decisions and patient prognosis. However, existing diagnostic techniques, such as magnetic resonance imaging (MRI), still face limitations, including a relatively delayed detection window and high examination costs, making it difficult to fully meet the clinical demand for rapid identification during the hyperacute phase. Optical molecular imaging based on abnormal changes in stroke-related endogenous metabolites provides a new approach for earlier lesion identification. Hydrogen polysulfides (H2Sn) and formaldehyde (FA) are two key metabolites that are synergistically elevated during the process of ischemic stroke, serving as potential dual targets for molecular imaging in the hyperacute phase. In this work, we obtained an H2Sn/FA cascade-responsive two-photon fluorescent probe, H2Sn-FA, which can sequentially recognize H2Sn and FA to trigger a cascade “turn-on” signal amplification, thereby improving the recognition specificity and imaging sensitivity in complex biological environments. Experimental results demonstrate that H2Sn-FA exhibits excellent selectivity and sensitivity toward the target molecules, along with low cytotoxicity. In a mouse model of ischemic stroke, H2Sn-FA enabled in vivo fluorescence imaging within 15 min, while MRI required up to 30 min to obtain an identifiable signal, saving 15 min for treatment decision. Furthermore, the probe allowed for the dynamic monitoring of secondary injury in the contralateral cerebral hemisphere induced by ischemia-reperfusion. This study establishes a H2Sn/FA cascade response strategy for hyperacute imaging of ischemic stroke; providing a potential tool for rapid molecular identification of stroke and laying a foundation for future clinical translation.