Find the papers that actually matter

PurposeLong-term survival after endovascular aneurysm repair (EVAR) for abdominal aortic aneurysm remains a clinical concern, particularly in elderly patients with comorbidities. This study aimed to compare different machine learning (ML) models that capture complex, nonlinear relationships among clinical variables to predict 5-year all-cause mortality following EVAR.MethodsWe retrospectively analyzed 142 patients who underwent elective EVAR between 2013 and 2018. Predictive models for 5-year mortality were developed using 3 supervised ML algorithms: Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Support Vector Classification (SVC). Each model was trained on the entire dataset and internally validated through 5-fold cross-validation. Model performance was evaluated using accuracy, sensitivity, specificity, precision, F1 score, and area under the curve (AUC) based on the training set and 5-fold cross-validation. Feature importance was assessed for RF and XGBoost.ResultsThe RF demonstrated the most consistent performance (training AUC 0.80; cross-validation AUC 0.77 ± 0.07). XGBoost achieved the highest training accuracy (0.85) but had lower cross-validation AUC (0.68 ± 0.05). SVC showed stable but modest performance. Key predictors identified by RF and XGBoost included poor nutritional status, octogenarian status, compromised immunity, and active cancer.ConclusionsTree-based ML models, especially RF, may effectively predict long-term survival after EVAR. Incorporating key clinical predictors into preoperative assessment may enhance risk stratification. Future studies should explore external validation and integration with time-to-event models such as Cox proportional hazards, to enhance prognostic accuracy.

In this research, key factors influencing breast cancer risk, a major global issue, are investigated, using machine learning (ML) and explainable AI, for racial differences. We used Breast Cancer Surveillance Consortium (BCSC) data, originally comprising 1.5 million unique combination records, from 6.7 million mammograms, collected between 2005 and 2017. Naïve Bayes, Logistic Regression, and Extreme Gradient Boosting models were applied to identify these key predictors. Variable importance and SHapley Additive exPlanations values were used to interpret models and identify most predictive factors. Analyses were stratified by six racial groups. History of biopsy (50.04%) and age group (25.85%) were the strongest predictors across all models and races. Menopausal status, breast density, and age at first childbirth were also important. White women had the highest overall incidences, particularly those over 65 (9.02 overall; 18.13 at age 65 + per 100,000), while Black women had higher rates in younger age groups (7.1 per 100,000 at age 18-29). Native American women showed higher rates in certain older age groups, whereas Asian/Pacific Islander and Other/Mixed groups had generally lower rates. ML and explainable AI applied to BCSC data identified key predictors and highlighted racial disparities among most predictive factors for breast cancer risk.

Respiratory motion is a long-standing challenge for lung stereotactic body radiotherapy (SBRT), particularly for centrally located lung tumors where increased toxicity demands more precise motion management during treatment. Current two-dimensional imaging approaches are insufficient for 3D tumor deformable motion tracking. In this study, we developed and evaluated a Surface-derived Three-dimensional (3D) AI-driven Real-time (STAR) imaging system by transforming a surface imaging system into a 3D real-time imaging solution. STAR integrated two key components: (1) prior-model-free spatiotemporal implicit neural representation (PMF-STINR): a machine-learning sub-system for pretreatment dynamic cone-beam CT (CBCT) reconstruction and motion modeling; and (2) surface-to-deformation network (Surf2DefNet): a deep-learning model that correlates intra-treatment body surface images with internal 3D anatomy and motion fields, trained based on the dynamic CBCT and motion model output of PMF-STINR. Specifically, PMF-STINR reconstructs a reference CBCT and solves an eigenvector-based motion model from a pretreatment CBCT, while Surf2DefNet predicts the motion eigen-vector weightings from surface images, enabling it to infer real-time CBCTs and motion vector fields (MVFs) using intra-treatment surface maps acquired later. We evaluated STAR imaging using both a digital extended cardiac torso (XCAT) phantom with regular and irregular motion patterns and ten patient datasets. The relative error (RE), center-of-mass error (COME), 95th percentile Hausdorff distance (HD95), Dice coefficients (DICE), and Pearson correlation coefficient (PCC) metrics between STAR images and the 'GT' were evaluated. For the XCAT phantom and the ten patients, the mean COME values are within 1 mm for all but one patient (1.3 mm). The RE values were consistently low, and the DICE and PCC values exceeded 0.89 in all cases. The HD95 are all within 2 mm except for one patient (2. 78 mm). These results demonstrate that the STAR imaging system can achieve accurate spatiotemporal reconstructions from surface images, providing CBCTs and MVFs for intra-treatment real-time image-guidance, and has the potential to improve safety and efficacy of SBRT for lung cancer.

This study aims to delineate the molecular mechanisms through which aristolochic acid (AA) exposure drives renal cell carcinoma (RCC) pathogenesis, leveraging an integrated machine learning (ML) and multi-omics approach to systematically identify and validate key targets linking AA to RCC development. By conducting differential expression analysis on multiple relevant datasets, we precisely screened for target genes closely associated with RCC. A multi-disciplinary approach was adopted, incorporating machine learning algorithms, network toxicology and molecular docking techniques, in order to elucidate the binding interactions between AA and its target proteins. Machine learning analysis identified eight core genes (HPD, SORD, CYP4F2, ERBB4, ALAD, PLAU, PYGL, KCNK5) as key regulatory factors. As core regulatory hub genes, molecular docking simulation results demonstrated predicted potential binding interactions between AA and target proteins. Cytological validation experiments have demonstrated that AA exhibits concentration gradient-dependent inhibition of renal cell viability, while simultaneously downregulating the expression of SORD. Clinical findings reveal that compared with normal kidney tissues, SORD is highly expressed in the cytoplasm of renal cancer tissues. This is the first study to systematically establish a comprehensive link between AA exposure and RCC tumorigenesis using an ML-guided multi-omics framework. We identified SORD as a key mechanistic player and demonstrated its functional and clinical relevance, providing insights into the molecular pathogenesis of AA-induced RCC. Our integrative approach offers a powerful strategy for elucidating environmental carcinogen-related mechanisms and identifying novel therapeutic targets.

Research on solute carrier transporters (SLCs) has become a well-established field, revealing the functional roles of these proteins in cancer biology. SLC transporters are transmembrane proteins that typically contain 7 to 12 transmembrane domains and mediate numerous essential functions, including transport of metabolites (such as glucose, amino acids, and lipids), signal transduction, immune cell interaction, and regulation of mitochondrial homeostasis. The expression of SLC transporters shows tissue, disease, and spatiotemporal specificity, and accumulating evidence indicates that SLC transporters are closely associated with pathological conditions, particularly tumor prognosis. Given their phylogenetic or species conservation, tumor tissue-specific expression, and potential to target specific pathways, SLC transporters hold promise as biomarkers for tumor diagnosis, treatment, and prognosis. SLC transporters play significant roles in tumor metabolism and tumor immunity, affecting not only tumor cells but also immune components of the tumor microenvironment, such as T cells, natural killer cells, and macrophages. In this review, we thoroughly examine the functions of SLCs in tumor metabolism immunity and tumor immunity, highlighting their impact on metabolic reprogramming of tumor cells and immune cell function within the tumor microenvironment. We also provide an overview of the SLC structure, SLC pharmacology, and recent advances in anticancer therapies targeting SLCs, particularly those with clinical efficacy. Finally, we discuss current challenges in the field of SLC transporters and propose future research directions.

The early diagnosis of ovarian cancer remains dissolved due to the lack of detection methods that are sensitive, rapid and accessible. To address this clinical challenge, this study developed an innovative diagnostic platform integrating a novel selenium nanoparticles (SeNPs) probe, dual-biomarker (CA125 and HE4) detection and intelligent analysis system. The detection limits for CA125 and HE4 were determined to be 35 U/mL and 100 pmol/L respectively, and no cross-reactivity was observed. Capitalizing on computer vision and AI prompt engineering technology, this study implemented smartphone-based image acquisition combined with AI-assisted analysis. The results of 360 clinical samples testing indicated that the coincidence rate of this kit with chemiluminescence method was over 95%, and the AUC of the combined detection of dual markers reached 0.914. By integrating the ROMA algorithm, the risk probability of ovarian cancer can be calculated in real time based on the test results and the patient's menopausal status, enabling quantitative risk assessment. This study provides a stable, highly specific, cost-effective and clinically decision supported point-of-care diagnostic solution for the early screening of ovarian cancer, promoting the development of POCT technology towards intelligence and precision.

The purpose of this study was to create and assess a new series of biphenyl urea derivatives as ferroptosis-inducing anticancer agents acting through GPX4 inhibition. Thirteen compounds were rationally designed, synthesized, and evaluated for their antiproliferative activity against the NCI-58 cancer cell panel. Several derivatives exhibited pronounced growth inhibition, particularly against SNB-75 CNS cancer cells, with low micromolar IC₅₀ values. Mechanistic investigations demonstrated that the most active compounds induce cancer cell death predominantly via ferroptosis rather than apoptosis, which was strongly associated with inhibition of the important lipid peroxidation regulator; glutathione peroxidase 4 (GPX4). Notably, compound 3e emerged as a lead molecule, showing sub-micromolar GPX4 inhibitory activity (IC₅₀ = 0.27 μM) and triggering ferroptosis through increased lipid peroxidation, generation of reactive oxygen species, and depletion of intracellular glutathione. Molecular docking and molecular dynamics simulations revealed persistent binding interactions inside the GPX4 active region, supporting these findings. Overall, this study identifies biphenyl urea derivatives as promising ferroptosis-inducing leads and provides a foundation for further optimization toward GPX4-targeted anticancer strategies.

Natural products have demonstrated significant potential in the treatment of many diseases. However, designing a diagnostic and therapeutic integrated system derived from natural products still poses significant challenges. This work proposes a drug delivery strategy based on natural products, which involves linking natural products to reactive moieties of tumor markers. Herein, a novel fluorescent probe (S-CES) for carboxylesterase 2 (CES2) imaging and therapy of colon tumor has been formulated and put to successful use both in vitro and in vivo. The probe was designed by combining trimethylacetate ester group with scopoletin, which itself can serve both as a tracer and drug in the treatment of colon tumor. S-CES exhibits high selectivity and sensitivity, low cytotoxicity, and excellent bio-compatibility, enabling the monitoring of endogenous CES2 activity in living cells. Moreover, it demonstrates favorable targeting specificity toward colon tumors. Consequently, this study offers a promising and robust strategy for the treatment of other diseases, while also augmenting the potential of natural products to exert a more significant role in clinical practice.

This study aimed to comprehensively investigate the anti-ovarian cancer (OC) efficacy of Yiyi Fuzi Baijiang Powder (YFBP), identify its key chemical constituents, and elucidate the underlying mechanisms of action. The chemical profile of YFBP was characterized using UHPLC-MS/MS. The potential mechanisms were predicted through integrated network pharmacology and bioinformatics analyses. The anti-tumor effects were validated in SK-OV-3 cells and a xenograft mouse model by performing CCK-8, flow cytometry, and TUNEL assays. The effects on the tumor inflammatory microenvironment and the JNK/c-Jun signaling pathway were assessed by ELISA, Western blotting, and qRT-PCR. UHPLC-MS/MS analysis identified 271 chemical constituents in YFBP. In vivo, YFBP significantly suppressed tumor growth (inhibition rate up to 52.4%) without systemic toxicity. It inhibited cell proliferation, induced apoptosis, and arrested the cell cycle in the S-phase. Mechanistically, YFBP ameliorated the inflammatory tumor microenvironment (TME) by reducing pro-inflammatory cytokine levels (TNF-α, IL-6, IL-1β) and concurrently activated the JNK/c-Jun signaling pathway. Rescue experiments confirmed that the JNK inhibitor SP600125 attenuated the anti-proliferative effects of YFBP. This study demonstrates that YFBP exerts significant anti-OC therapeutic effects by modulating the inflammatory TME and activating the JNK/c-Jun pathway. Our findings provide a pharmacological basis for the traditional use of YFBP and highlight its potential as a promising candidate phytomedicine for OC therapy.

Metalloproteinases (MPs) such as a-disintegrin and metalloproteinase-10 (ADAM-10) are key drivers of extracellular matrix remodeling during tumor progression, yet MP-based liquid biopsy tests have not reached clinical utility. Here, we show that active ADAM-10 is selectively enriched on the surface of circulating extracellular vesicles (EVs) in the plasma of colorectal cancer patients. Our findings further suggest ADAM-10+ EVs are locally enriched in dense pre-metastatic tumor extracellular matrices and subsequently accumulate in blood post-metastasis. To capture these unique signatures of disease progression, an ADAM-10 activity assay is integrated with a novel size-sensitive Immuno-Janus Particle affinity assay for characterizing ADAM-10+ EVs in untreated plasma. In a 43-patient colorectal cancer cohort, this multimodal platform distinguishes healthy, pre-metastatic, and metastatic states with 95% overall accuracy. When combined with lipidomics as a third modality, the platform correctly determines 97.4% cancer stage accuracy, with only one misclassification. This study establishes a multimodal EV-based activity/affinity assay as a robust framework for liquid biopsy, providing accurate cancer staging, improved prognostics, and offering a potential platform for pan-disease diagnostics.