Background/Aims: Liver transplantation (LT) following total hepatectomy is a life-saving treatment for hepatocellular carcinoma (HCC). The HCC recurrence after LT hinders the effectiveness of the procedure. The objective of this study is to develop a pre-operative risk stratification model based on a liquid biopsy. Methods: We conducted a comprehensive multi-omics study of 260 HCC patients from three centers, including clinical data, low-coverage whole-genome sequencing of cell-free DNA (cfDNA) from plasma, as well as whole-exome, single-nucleus RNA, and spatial transcriptomics from matched tumor and non-tumor tissues. Results: We identified cfDNA-derived copy number alteration (CNA) signatures associated with post-transplant recurrence. By integrating cfDNA-derived CNA profiles with single-cell transcriptomic data, we traced recurrence-associated cfDNA to a distinct subpopulation of malignant cells within the primary tumor. These cells were embedded in a pro-metastatic microenvironment of specialized endothelial subtypes and cancer-associated fibroblasts. Notably, most recurrence-associated lesions were detectable in cfDNA prior to liver transplantation (LT). Building on these insights, we developed the ZJU Criteria based on CNA fragments and tumor markers, a pre-LT risk prediction tool that integrates conventional clinical factors with cfDNA-derived CNA signatures, and validated it using internal and independent external cohorts. Conclusion Our findings suggest that post-transplant recurrence commonly originates from advanced subclones that emerge late during tumor evolution. The ZJU Criteria provides an accurate, non-invasive strategy that significantly improves pre-LT risk stratification and clinical decision-making for patients with HCC.

Aim To explore the compatibility and po-tential mechanism of effective components of Biejia-Ezhu against triple negative breast cancer(TNBC)and verify it by experiments.Methods Effective compo-nents and targets of Biejia-Ezhu were obtained by TC-MSP and Swiss Target Prediction.Disease targets of TNBC were obtained from OMMI and GeneCards data-bases.The PPI network was constructed using STRING database.GO and KEGG path enrichment analysis was performed using DAVID database.Cytoscape3.9.1 software was used to construct the"drug-component-target-disease"network,screen key targets and compo-nents for molecular docking,and further verify the com-patibility of key components and targets in vitro.Re-sults ① A total of 71 effective components were iden-tified in the Biejia-Ezhu drug pair.There were 146 drug targets associated with the disease.A total of 113 signaling pathways were identified by KEGG analysis.The 71 potential active components of Biejia-Ezhu mainly acted on key targets such as mTORC1,ULK1,TNF,EGFR,ESR1,STAT3,HIF1A,and PTGS2.Mo-lecular docking results showed that glycine and curcu-min were the key active components of Biejia-Ezhu,and both had strong docking activity against key target proteins mTORC1 and ULK1.②The results of in vitro experiment showed that glycine combined with curcu-min significantly inhibited the proliferation and clonal formation ability of TNBC cells(P＜0.05),up-regula-ted the expression of autophagy marker LC3 Ⅱ/Ⅰ,down-regulated the expression of EGFR,down-regula-ted the expression of pathway protein mTORC1,p-mTOR,p-ULK1,and promoted the expression of path-way protein ULK1(P＜0.05).Conclusion The key component of Biejia-Ezhu against triple-negative breast cancer is glycine-curcumin,the mechanism of which may be related to the regulation of the mTORC1/ULK1 signaling pathway to promote autophagy.

With the rapid rise of large language models(LLM),the natural language generation capabilities of deep learning have demonstrated significant value in the medical field.However,the"closed nature"of model parameters makes them prone to generating"hallucinations",making it difficult to provide accurate answers to the latest knowledge,and the reasoning process lacks transparency and traceability.Retrieval-augmented generation(RAG)technology addresses these issues by actively connecting external information sources such as document databases and knowledge graphs during the generation process.This significantly reduces the dependence of LLM on outdated training data and introduces verifiable evidence and real-time knowledge updates into their responses.In the medical field,RAG technology effectively addresses the high-accuracy and traceability requirements of literature retrieval and clinical decision support.It is widely applied in areas such as drug discovery,pharmacovigilance,and the diagnosis and treatment of rare diseases.By integrating emerging technologies such as reinforcement learning,multimodal processing,and compliant privacy protection,RAG technology is evolving towards a more open and highly customizable direction,providing innovative intelligent solutions for medical information retrieval and decision-making support.

Objective To validate the diagnostic performance and risk stratification ability of an AI-based recognition system(PE-AI)for pulmonary embolism(PE)using computed tomography pulmonary angiography(CTPA)so as to analyze its diagnostic value in clinical practice.Methods A total of 416 patients with suspected PE who underwent CTPA from January 1,2023 to December 10,2023 at our hospital were included in this study.Two junior radiologists and PE-AI separately detected and diagnosed emboli in the collected cases by double-blind method,and recorded the diagnosis time respectively.Three senior radiologists reviewing with clinical follow-up results were used as the gold standard in this study.Diagnostic performance was evaluated by using the receiver operating characteristic(ROC)curve analysis and Delong-t test.For positive cases,the pulmonary artery obstruction index(PAOI)calculated by AI and manually were collected respectively and consistency analysis was performed.Results The area under the curve(AUC)of PE-AI,manual and combined diagnosis was 85.6％,90.8％and 95.1％,respectively,which differed significantly(P＜0.05).The reading time of PE-AI[(0.16±0.07)min]was significantly lower than the time of manual[(4.42±1.85)min,P＜0.001]and combined diagnosis[(4.58±1.84)min,P＜0.001].The PAOI measured by PE-AI and manually had high consistency(intraclass correlation efficient,ICC=0.80)in the subgroup analysis of confirmed cases.Conclusion AI can quickly identify pulmonary artery emboli in a short time and assist radiologists to improve diagnostic efficiency.At the same time,through the intelligent detection of PAOI,it is helpful for the risk stratification of patients with PE and optimizing the diagnosis and treatment pathway for pulmonary embolism.

Objective:To analyze the clinical characteristics, treatments, and prognosis of patients with ovarian juvenile granulosa cell tumor (JGCT).Methods:Clinical and pathological data, and follow-up information of 34 patients diagnosed with JGCT from 2000 to 2021 were collected from the surveillance, epidemiology, and end results (SEER) database. A retrospective analysis was conducted to summarize the patients′ clinical and pathological characteristics, treatments, and prognosis. Propensity score matching (PSM) was used to match the JGCT cases with adult granulosa cell tumor (AGCT) cases in SEER database. A total of 96 patients with ovarian granulosa cell tumor (OGCT), including 32 cases of JGCT and 64 cases of AGCT, were enrolled in a matched cohort analysis. Univariate and multivariate Cox regression analysis were performed on the matched cohort to explore the risk factors for overall survival. Kaplan-Meier curves and the log-rank test were used to compare the survival outcomes between JGCT and AGCT.Results:(1) The median age at diagnosis for the 34 JGCT patients was 19.5 years (ranged: 1-48 years), with 3 patients aged ≤10 years, 16 patients aged 11-20 years, 11 patients aged 21-30 years, and 4 patients aged >30 years. Tumors originated unilaterally in 33 patients, with only 1 case originating bilaterally. The maximum tumor diameter was recorded in 26 patients, with a median size of 12.4 cm (ranged: 3.5-40.0 cm). According to the 2014 International Federation of Gynecology and Obstetrics (FIGO) staging system, 19 patients were diagnosed with stage Ⅰ (including 10 cases with stage Ⅰa and 9 cases with stage Ⅰc), 4 patients with stage Ⅱ, 8 patients with stage Ⅲ, and 3 patients with stage Ⅳ. Two patients did not undergo surgery for the resection of lesions. Stage Ⅰ patients (15/19) underwent fertility-sparing surgery, while stage Ⅱ-Ⅲ patients underwent either fertility-sparing surgery or cytoreductive surgery (6 cases each). Stage Ⅳ patients underwent cytoreductive surgery (2 cases). Lymph node dissection was performed in 10 patients, among which only 1 patient with positive lymph nodes metastasis. None of the 34 patients received radiotherapy, while 18 patients received adjuvant chemotherapy (included neoadjuvant chemotherapy and postoperative adjuvant chemotherapy). The proportion of stage Ⅰ patients receiving adjuvant chemotherapy was relatively low, with only 4 out of 19 patients (including 2 out of 10 cases for stage Ⅰa and 2 out of 9 cases for stage Ⅰc). The proportions of patients receiving adjuvant chemotherapy for stages Ⅱ, Ⅲ and Ⅳ were 3 out of 4 cases, 8 out of 8 cases, and 3 out of 3 cases, respectively. The follow-up ended in December 2021, with 20 patients alive and 14 dead. The survival rate for ovarian JGCT patients was 59% (20/34). Among them, the survival rate for stage Ⅰ patients was 16/19, while for stage Ⅱ-Ⅳ patients, it was 4/15; there was a statistically significant difference ( P=0.002). Among stage Ⅱ-Ⅲ patients, the survival rate at the end of follow-up was 1/6 for those who underwent fertility-sparing surgery, compared to 3/6 for those who underwent cytoreductive surgery ( P=0.546). (2) For the 96 OGCT patients after matching using the PSM method, 64 ovarian AGCT patients had 5 deaths and 59 survivors during the follow-up period, the survival rate was 92% (59/64) at the end of follow-up. In contrast, among the 32 ovarian JGCT patients, 13 died and 19 survived, resulting in a survival rate of 59% (19/32) at the end of follow-up, which was statistically significant difference for the AGCT group ( P<0.001). Univariate Cox analysis revealed that histology, extent of surgery, chemotherapy, postoperative tumor residual status, and stage all significantly affected the survival outcomes of OGCT patients (all P<0.05). Multivariate Cox analysis revealed that variables with significant statistical differences were histology and stage. The median survival time for JGCT patients was 126 months, while AGCT patients median survival time was not reached with a statistically significant between the two groups ( P<0.001). Conclusions:Ovarian JGCT predominantly occur in adolescents and young women. Lymph node metastasis is relatively rare, and treatment primarily involves surgery and adjuvant chemotherapy. Most ovarian JGCT patients are diagnosed at stage Ⅰ, with a favorable prognosis. Fertility-preserving surgery is recommended, involving salpingo-oophorectomy on the affected side plus comprehensive staging surgery, or a second surgery to achieve comprehensive staging. For stage Ⅱ-Ⅳ ovarian JGCT patients, the prognosis is relatively poor, and fertility-preserving surgery should be considered with caution. The prognosis of ovarian JGCT patients is worse than that of ovarian AGCT patients.

Aim To explore the compatibility and po-tential mechanism of effective components of Biejia-Ezhu against triple negative breast cancer(TNBC)and verify it by experiments.Methods Effective compo-nents and targets of Biejia-Ezhu were obtained by TC-MSP and Swiss Target Prediction.Disease targets of TNBC were obtained from OMMI and GeneCards data-bases.The PPI network was constructed using STRING database.GO and KEGG path enrichment analysis was performed using DAVID database.Cytoscape3.9.1 software was used to construct the"drug-component-target-disease"network,screen key targets and compo-nents for molecular docking,and further verify the com-patibility of key components and targets in vitro.Re-sults ① A total of 71 effective components were iden-tified in the Biejia-Ezhu drug pair.There were 146 drug targets associated with the disease.A total of 113 signaling pathways were identified by KEGG analysis.The 71 potential active components of Biejia-Ezhu mainly acted on key targets such as mTORC1,ULK1,TNF,EGFR,ESR1,STAT3,HIF1A,and PTGS2.Mo-lecular docking results showed that glycine and curcu-min were the key active components of Biejia-Ezhu,and both had strong docking activity against key target proteins mTORC1 and ULK1.②The results of in vitro experiment showed that glycine combined with curcu-min significantly inhibited the proliferation and clonal formation ability of TNBC cells(P＜0.05),up-regula-ted the expression of autophagy marker LC3 Ⅱ/Ⅰ,down-regulated the expression of EGFR,down-regula-ted the expression of pathway protein mTORC1,p-mTOR,p-ULK1,and promoted the expression of path-way protein ULK1(P＜0.05).Conclusion The key component of Biejia-Ezhu against triple-negative breast cancer is glycine-curcumin,the mechanism of which may be related to the regulation of the mTORC1/ULK1 signaling pathway to promote autophagy.

Objective:To investigate the predictive value of serum neutrophil gelatinase-associated lipocalin(NGAL),calcitonin gene-related peptide(CGRP),and neutrophil-to-lymphocyte ratio(NLR)for the prognostic regression of elderly patients with stroke complicated with pulmonary infection(SCPI).Methods:This study was a retrospective single-center study,149 elderly patients with SCPI who were admitted to Inner Mongolia Baogang Hospital from June 2020 to June 2024 were selected,they were divided into poor prognosis group(n=56)and good prognosis group(n=93)according to the prognosis.Baseline data and laboratory test indicators were collected,and NLR was calculated.Serum NGAL and CGRP levels were measured by ELISA.Influencing factors of poor prognosis of elderly patients with SCPI were analyzed by Multivariate logistic regression.Predicts value was analyzed by Receiver operating characteristic(ROC)curve.Results:Compared with good prognosis group,the poor prognosis group had higher of aged ≥ 70 years,incidence of hemorrhagic stroke,serum creatinine,white blood cell count,national institute of health stroke scale(NIHSS),platelet count,C-reactive protein,NGAL,and NLR levels,longer nerosurgery intensive care unit(NICU)stay,and lower CGRP levels(P＜0.05).Higher CGRP level was an independent protective factor of poor prognosis of elderly patients with SCPI(OR＜1,P＜0.05).Age ≥ 70 years,hemorrhagic stroke,longer NICU stay,higher NIHSS score,higher NGAL level and higher NLR were independent risk factors of poor prognosis of elderly patients with SCPI(OR＞1,P＜0.05).The area under the curve(AUC)for predicting the prognostic regression of elderly patients with SCPI used NGAL,CGRP,and NLR alone or in combination was 0.777,0.771,0.786,and 0.927,respectively,with the combination of three factors showed the highest predictive power(P＜0.05).Conclusion:Age ≥70 years,hemorrhagic stroke,longer NICU stay,higher NIHSS score,higher NGAL level and higher NLR are independent risk factors of poor prognosis of elderly patients with SCPI,while higher CGRP level is an independent protective factor.The combination detection of NGAL,CGRP and NLR can improve the predictive value of prognostic regression in elderly patients with SCPI.

Objective:To investigate the predictive value of serum neutrophil gelatinase-associated lipocalin(NGAL),calcitonin gene-related peptide(CGRP),and neutrophil-to-lymphocyte ratio(NLR)for the prognostic regression of elderly patients with stroke complicated with pulmonary infection(SCPI).Methods:This study was a retrospective single-center study,149 elderly patients with SCPI who were admitted to Inner Mongolia Baogang Hospital from June 2020 to June 2024 were selected,they were divided into poor prognosis group(n=56)and good prognosis group(n=93)according to the prognosis.Baseline data and laboratory test indicators were collected,and NLR was calculated.Serum NGAL and CGRP levels were measured by ELISA.Influencing factors of poor prognosis of elderly patients with SCPI were analyzed by Multivariate logistic regression.Predicts value was analyzed by Receiver operating characteristic(ROC)curve.Results:Compared with good prognosis group,the poor prognosis group had higher of aged ≥ 70 years,incidence of hemorrhagic stroke,serum creatinine,white blood cell count,national institute of health stroke scale(NIHSS),platelet count,C-reactive protein,NGAL,and NLR levels,longer nerosurgery intensive care unit(NICU)stay,and lower CGRP levels(P＜0.05).Higher CGRP level was an independent protective factor of poor prognosis of elderly patients with SCPI(OR＜1,P＜0.05).Age ≥ 70 years,hemorrhagic stroke,longer NICU stay,higher NIHSS score,higher NGAL level and higher NLR were independent risk factors of poor prognosis of elderly patients with SCPI(OR＞1,P＜0.05).The area under the curve(AUC)for predicting the prognostic regression of elderly patients with SCPI used NGAL,CGRP,and NLR alone or in combination was 0.777,0.771,0.786,and 0.927,respectively,with the combination of three factors showed the highest predictive power(P＜0.05).Conclusion:Age ≥70 years,hemorrhagic stroke,longer NICU stay,higher NIHSS score,higher NGAL level and higher NLR are independent risk factors of poor prognosis of elderly patients with SCPI,while higher CGRP level is an independent protective factor.The combination detection of NGAL,CGRP and NLR can improve the predictive value of prognostic regression in elderly patients with SCPI.

Previous studies have shown that the diversity and composition of intestinal microbiota are related to the effect of tumor immunotherapy,but the mechanism of intestinal microbiota affecting tumor immunotherapy resistance has rarely been sum-marized.This article not only expounds the current clinical sta-tus of tumor immunotherapy resistance,but also summarizes the correlation and regulatory mechanism between the composition and homeostasis of intestinal microbiota and drug resistance to different types of tumor immunotherapy,so as to provide a refer-ence for the study of potential targets for improving tumor immu-notherapy resistance based on intestinal microbiota.

With the rapid rise of large language models(LLM),the natural language generation capabilities of deep learning have demonstrated significant value in the medical field.However,the"closed nature"of model parameters makes them prone to generating"hallucinations",making it difficult to provide accurate answers to the latest knowledge,and the reasoning process lacks transparency and traceability.Retrieval-augmented generation(RAG)technology addresses these issues by actively connecting external information sources such as document databases and knowledge graphs during the generation process.This significantly reduces the dependence of LLM on outdated training data and introduces verifiable evidence and real-time knowledge updates into their responses.In the medical field,RAG technology effectively addresses the high-accuracy and traceability requirements of literature retrieval and clinical decision support.It is widely applied in areas such as drug discovery,pharmacovigilance,and the diagnosis and treatment of rare diseases.By integrating emerging technologies such as reinforcement learning,multimodal processing,and compliant privacy protection,RAG technology is evolving towards a more open and highly customizable direction,providing innovative intelligent solutions for medical information retrieval and decision-making support.