Objective: To construct a nano delivery system based on sonodynamic therapy to overcome cisplatin resistance of ovarian cancer by inducing pyroptosis.

Methods: A human cisplatin-resistant ovarian cancer cell line (SKOV-3/DDP) was successfully induced using a stepwise concentration gradient induction method. A composite nanoparticles (BTO/ZIF) with a core of barium titanate (BaTiO₃, BTO) and a shell of zeolitic imidazolate framework-8 (ZIF-8) was constructed. The surface of BTO/ZIF was further coated with DSPE-PEG₁₀₀₀-Pt (Ⅳ) and DSPE-PEG₁₀₀₀-AF to prepare the targeted nanoparticle BTO/ZIF@AF/Pt. The morphology, size, structure, hydrodynamic diameter, and colloidal stability (Zeta potential) of BTO/ZIF@AF/Pt nanoparticles were characterized using transmission electron microscopy, scanning electron microscopy, and Malvern particle size analyzer. The reactive oxygen species (ROS) generation capability of BTO/ZIF composite nanoparticles was assessed by ultraviolet-visible (UV) spectrophotometry and DCFH-DA probe assay. Cellular uptake of BTO/ZIF@AF/Pt nanoparticles by SKOV-3/DDP cells was verified using confocal laser scanning microscopy. Under ultrasound irradiation, the effects of BTO/ZIF@AF/Pt nanoparticles on the expression of pyroptosis-related proteins and sensitivity to cisplatin chemotherapy were examined by Western blotting.

Results: In this study, a human cisplatin-resistant ovarian cancer cell line (SKOV-3/DDP) was successfully induced, and a targeted delivery nanoparticle using BTO and metal-organic framework (MOF) composite materials as the carrier for cisplatin was also successfully constructed. Under ultrasound stimulation, BTO/ZIF@AF/Pt nanoparticles effectively generated ROS and induced pyroptosis in ovarian cancer SKOV-3/DDP cells.

Conclusion: Ultrasound-triggered ROS generation from BTO/ZIF@AF/Pt nanoparticles, combined with Zn^2＋ release, synergistically promotes ROS-induced pyroptosis in ovarian cancer SKOV-3/DDP cells, thereby effectively overcoming cisplatin resistance.

Objective: To construct a unmethylated CpG (cytosine-phosphate-guanine) nano delivery system based on the cationic polymer carrier polyethylenimine (PEI), termed CpG-PEI, and to evaluate its efficiency in promoting intracellular delivery of CpG, tis ability to enhance anti-tumor immunotherapy, and its biosafety.

Methods: CpG-PEI nanoparticles were prepared through the electrostatic self-assembly method. The synthesis morphology, particle size, and zeta potential of CpG-PEI nanoparticles were characterized by agarose gel electrophoresis, field emission scanning electron microscopy, and dynamic light scattering (DLS); The uptake efficiency of CpG-PEI nanoparticles by RAW264.7 macrophages was assessed using flow cytometry and laser scanning confocal microscopy. The impact CpG-PEI nanoparticles on macrophage polarization was evaluated by detecting M1/M2 phenotypic markers on the surface of RAW264.7 macrophages via flow cytometry. The cytotoxicity and apoptosis-inducing effects of CpG-PEI nanoparticles were evaluated using the CCK-8 assay and Annexin V-FITC/PI double staining method to verify its biosafety.

Results: CpG-PEI nanoparticles with a particle size of 200−400 nm were successfully prepared, its electronegativity gradually changes towards a positive charge direction as the proportion of PEI in the synthesis system increases; CpG-PEI nanoparticles significantly enhanced the intracellular uptake of CpG by RAW264.7 macrophages and effectively induced macrophage polarization towards anti-tumor M1 phenotype. By adjusting the proportion of PEI in the CpG-PEI synthesis system, the cytotoxicity of the nanoparticles could be significantly reduced while maintaining efficient delivery and immune activation.

Conclusion: This study successfully developed a safe and effective CpG-PEI nanodelivery system, which enables efficient intracellular delivery of CpG in immune cells and induces polarization of RAW264.7 macrophages towards the anti-tumor M1 phenotype, providing a new delivery strategy for nucleic acid adjuvant-based tumor immunotherapy.

Objective: To investigate the effect of overexpression of exogenous ETS translocation variant 2 (ETV2) gene on the tube-forming ability of human umbilical vein endothelial cells (HUVECs) and to construct an endothelial cell-tumor organoid co-culture model capable of adapting to tumor growth.

Methods: A doxycycline-inducible Tet-On plasmid system was used to regulate the expression level of ETV2 gene, which was transfected into HUVECs via a lentiviral vector to establish a stable transfection HUVEC cell line (named ETV2-EC) with ETV2 expression modulated by doxycycline concentration. The tube-forming ability of ETV2-EC cells induced with different concentrations of doxycycline was assessed through tube formation assay. Subsequently, ETV2-EC cells were co-cultured with breast cancer organoids, and the optimal co-culture medium for the co-culture model was determined via Calcein-AM staining and CCK-8 assay. The viability of endothelial cells and their interactions with tumor organoids in the co-culture model were observed under an optical microscope. Finally, the effect of ETV2 overexpression on the transcriptome profile of HUVECs was analyzed by RNA-seq.

Results: Induction with 10 μmol/L doxycycline significantly enhanced the tube-forming ability and tubular structure stability of ETV2-EC cells. Both endothelial cells and tumor organoids could stably survive in the breast cancer organoid culture medium containing fibroblast growth factor 2 (FGF2) and heparin. Compared with control HUVECs, ETV2-EC cells induced with 10 μmol/L doxycycline showed longer survival time, larger tubular structure area, and greater stability when co-cultured with tumor organoids. RNA-seq analysis revealed that overexpression of ETV2 gene markedly altered the transcriptomic profile of HUVECs, and the differentially expressed genes in ETV2-EC cells were mainly associated with endothelial cell proliferation, migration, and adaptation to the tumor microenvironment.

Conclusion: Overexpression of ETV2 gene in mature endothelial cells may endow them with enhanced tube-forming ability by reactivating the suppressed genes associated with lumen formation and angiogenesis. Moreover, under co-culture conditions with tumor organoids, overexpression of ETV2 gene significantly improves the survival ability of endothelial cells and their adaptation to the tumor microenvironment.

Objective: To investigate the effect of serum bile acids, especially taurochenodeoxycholic acid (TCDCA), on gemcitabine resistance and survival prognosis in patients with gallbladder cancer.

Methods: Fresh gallbladder cancer tissues from 20 patients with gallbladder cancer were selected to construct mini patient derived xenograft (mini-PDX) models and patient-derived organoids. These patients were divided into a gemcitabine sensitive group and a gemcitabine resistant group using mini-PDX drug sensitivity testing. To detect serum bile acids in gemcitabine-sensitive and gemcitabine-resistant patients with gallbladder cancer using high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS), and compare the differences between the two groups. Using Kaplan-Meier method and COX proportional hazards model to assess the association between serum TCDCA levels and overall survival of gallbladder cancer patients, and conducting subgroup analysis of gallbladder cancer patients based on whether they received gemcitabine treatment. In the GBC-SD cell line and gallbladder cancer organoid models, the effects of TCDCA on gemcitabine resistance were assessed using CCK-8 cytotoxicity assay, colony formation assay, and organoid drug sensitivity test.

Results: Compared with gemcitabine-sensitive group, multiple conjugated bile acids were elevated in serum from gemcitabine-resistant patients, with TCDCA showing the most pronounced increase (P=3.3×10⁻⁵). In the clinical cohort, patients with high TCDCA had significantly shorter overall survival than those with low TCDCA [hazard ratio (HR)=1.75, 95% confidence interval (CI): 1.18-2.59; P=0.009]. Subgroup analysis revealed that this adverse prognostic association was confined to patients treated with gemcitabine (HR=2.07, 95% CI: 1.23-3.47; P=0.005), whereas no significant association in those not receiving gemcitabine (P=0.338). In vitro functional assays, TCDCA treatment significantly increased the IC₅₀ of gemcitabine in GBC-SD cells, and markedly enhanced the colony forming ability of GBC-SD cells under chemotherapeutic pressure. In gallbladder cancer organoids, TCDCA similarly attenuated gemcitabine-induced growth inhibition.

Conclusion: Serum TCDCA is a potential adverse prognostic biomarker for patients with gallbladder cancer receiving gemcitabine chemotherapy. TCDCA can directly induce gemcitabine resistance in GBC-SD cells and gallbladder cancer organoids, suggesting that targeting the bile acid metabolic microenvironment may provide a feasible therapeutic strategy to reverse gemcitabine chemotherapy resistance in gallbladder cancer.

Lung cancer is one of the most common and deadly malignant tumors worldwide, with its significant tumor heterogeneity and drug resistance being major obstacles to achieving precision medicine. Patient-derived tumor organoids, which can faithfully recapitulate the genetic and pathological features of the original tumor, serve as a bridge linking basic research and clinical practice, and show great promise in advancing precision therapy for lung cancer. This paper systematically reviews the construction strategies and technical advantages of lung cancer organoid models, and points out that the key to future development lies in establishing standardized culture systems, conducting multicenter clinical validation, and deepening the integration of multi-omics data, with the aim of promoting the application of lung cancer organoids in drug development and individualized precision therapy.

Objective: To establish a predictive model of platelet transfusion efficacy in patients with myelodysplastic syndrome (MDS) based on machine learning algorithms, and to validate its predictive performance and clinical utility in a patient cohort.

Methods: A total of 160 MDS patients in Yuncheng Central Hospital Affiliated to Shanxi Medical University from January 2021 to December 2024 were enrolled as the training set and divided into an ineffective group and an effective group according to platelet transfusion efficacy. Characteristic variables for constructing the prediction model of platelet transfusion efficacy in MDS patients were sequentially screened using least absolute shrinkage and selection operator (LASSO) regression analysis and logistic regression analysis. Four machine learning algorithms, namely extreme gradient boosting (XGBoost), decision tree, random forest and logistic regression, were respectively employed to construct the prediction models. Predictive performance of each model was evaluated using sensitivity, specificity, the area under the curve （AUC） of receiver operating characteristic (ROC) and Youden index. The decision curve analysis was further used to evaluate the clinical practicability of the optimal model. In addition, 73 patients with MDS in Yuncheng Central Hospital Affiliated to Shanxi Medical University and the Second Hospital of Shanxi Medical University from January to May 2025 were selected as the time validation set to evaluate the clinical generalizability of the optimal model.

Results: Among the 160 patients with MDS in the training set, the rate of platelet transfusion inefficiency was 34.38% (55/160). The proportions of fever, splenomegaly, platelet antibody positive and platelet transfusion times ≥5 times, as well as the levels of serum interleukin (IL)-1β and IL-8 in the ineffective group were higher than those in the effective group (all P < 0.05). Logistic regression analysis showed that splenomegaly, platelet antibody positive, platelet transfusion times and the levels of serum IL-1β and IL-8 were independent risk factors for ineffective platelet transfusion in MDS patients (all P < 0.05). The AUC values of XGBoost model were 0.946 [95% confidence interval (CI): 0.899-0.975] in the training set and 0.947 (95% CI: 0.868-0.986) in the validation set, which were both significantly higher than those of the random forest model [0.871 (95% CI: 0.809-0.919) and 0.830 (95% CI: 0.723-0.907)], decision tree model [0.856 (95% CI: 0.792-0.907) and 0.814 (95% CI: 0.705-0.895)] and logistic regression model [0.849 (95% CI: 0.784-0.901) and 0.804 (95% CI: 0.695-0.888)] (all P < 0.05). Additionally, in both the training and validation sets, the XGBoost model demonstrated the sensitivities of 90.91% and 91.67%, and the specificities of 89.52% and 91.84%, which were both significantly higher than those of the other three predictive models (both P < 0.05). The XGBoost model ranked the factors that increase the risk of ineffective platelet transfusion in MDS patients in order of importance, namely platelet antibody positive, platelet transfusion frequency, serum IL-8 level， serum IL-1β level and splenomegaly. The decision curve showed that when the threshold probability ranged from 0.01 to 0.99, applying the XGBoost model to predict the efficacy of platelet transfusion in MDS patients consistently demonstrates that the clinical benefits derived from correct interventions outweigh the losses caused by misjudgments, which has good clinical practical value.

Conclusion: The efficacy of platelet transfusion therapy for MDS patients is not ideal. The XGBoost model, constructed based on five clinical indicators including platelet antibody positivity, platelet transfusion frequency, serum IL-8 level, serum IL-1β level and splenomegaly, has the best comprehensive predictive performance for the platelet transfusion effect in MDS patients and has good clinical practical value. It can provide references for clinical adjustment, optimization of treatment plans and improvement of platelet transfusion efficiency.

Primary pulmonary lymphoma (PPL) is relatively rare malignancy in clinical practice, whose imaging and pathological phenotypes lack specificity. Traditional diagnosis primarily relies on clinicians' personal experience, which can easily lead to misdiagnosis and missed diagnoses. Machine learning (ML), as a core technology of artificial intelligence (AI), provides a new approach to improving the diagnostic efficacy of PPL. This article reviews the latest advances in ML applications for PPL diagnosis. In the field of radiomics, ML facilitates the extraction of high-throughput texture features from CT and positron emission tomography (PET) imaging to construct quantitative models, thereby significantly improving the diagnostic accuracy for differentiating PPL from lung cancer, organizing pneumonia, and other diseases. In the field of pathomics, deep learning techniques have enabled automated classification of lymphoma subtypes and objective immunohistochemical analysis. Additionally, genomics intergrated with ML helps elucidate the molecular characteristics of PPL to support its precise classification. Although the application of ML in PPL diagnosis still face challenges such as limited sample sizes, insufficient standardization and poor model interpretability, advancements in multimodal data integration, federated learning and interpretable AI are expected to propel the diagnosis and treatment of PPL toward intelligent and precise development.

Retroperitoneal soft tissue sarcoma (RPSTS) is a rare malignant tumor, and its diagnosis and treatment face significant challenges due to the high heterogeneity of the tumor. Currently, imaging examinations, such as CT and MRI, are key tools for diagnosing RPSTS. However, traditional methods have limitations in achieving precise evaluation. In recent years, machine learning technologies have shown considerable potential to further enhance the diagnosis and treatment of RPSTS. Machine learning has achieved remarkable results in subtype classification, grading, treatment monitoring, and prognostic prediction for RPSTS. For example, it can effectively distinguish different subtypes of RPSTS, such as well-differentiated liposarcoma and dedifferentiated liposarcoma, and predict important indicators, including tumor grade, treatment response, and the risk of distant metastasis. Although machine learning has made significant progress in image data processing and model optimization, the subjectivity of manual segmentation of regions of interest (ROI), the constraints of small sample data, and the “black-box” nature of deep learning models remain critical bottlenecks restricting its clinical implementation. In the future, through large-scale multicenter studies, the development of automated segmentation technologies, and improvements in model interpretability, machine learning is expected to gradually transform into a reliable clinical tool, which provides more accurate diagnostic and treatment to RPSTS patients and ultimately delivering real benefits to them.

Pancreatic cancer, a highly malignant tumor of digestive system, is characterized by challenges in early detection and limited therapeutic efficacy. Currently, artificial intelligence (AI) technologies—such as machine learning, deep learning, and natural language processing (NLP)—are providing innovative strategies to address key bottlenecks in the diagnosis and treatment of pancreatic cancer. Based on databases such as PubMed, Web of Science, and CNKI, this study conducted literature retrieval and screening, initially obtaining 128 articles related to AI-powered pancreatic cancer diagnosis and treatment. After in-depth review, 38 journal papers and research reports were ultimately included, covering key aspects such as AI-based disease risk prediction models, intelligent screening of medical imaging and pathological assistance diagnosis, personalized treatment strategy recommendations, efficacy monitoring, and recurrence prediction. The study systematically reviews the integration and application of AI technologies throughout the entire process of pancreatic cancer diagnosis and treatment. By conducting a meta-analysis of current practices and evidence regardin AI in pancreatic cancer diagnosis and treatment, this paper aims to provide academic references for promoting the effective translation of AI technologies into clinical practice for pancreatic cancer.