Renal Failure最新文献

Introduction: Natural language processing (NLP) has been used to analyze unstructured imaging report data, yet its application in identifying chronic kidney disease (CKD) features from kidney ultrasound reports remains unexplored.

Methods: In a single-center pilot study, we analyzed 1,068 kidney ultrasound reports using NLP techniques. To identify kidney echogenicity as either "normal" or "increased," we used two methods: one that looks at individual words and another that analyzes full sentences. Kidney length was identified as "small" if its length was below the 10th percentile. Nephrologists reviewed 100 randomly selected reports to create the reference standard (ground truth) for initial model training followed by model validation on an independent set of 100 reports.

Results: The word-level NLP model outperformed the sentence-level approach in classifying increased echogenicity (accuracy: 0.96 vs. 0.89 for the left kidney; 0.97 vs. 0.92 for the right kidney). This model was then applied to the full dataset to assess associations with CKD. Multivariable logistic regression identified bilaterally increased echogenicity as the strongest predictor of CKD (odds ratio [OR] = 7.642, 95% confidence interval [CI]: 4.887-11.949; p < 0.0001), followed by bilaterally small kidneys (OR = 4.981 [1.522, 16.300]; p = 0.008). Among individuals without CKD, those with bilaterally increased echogenicity had significantly lower kidney function than those with normal echogenicity.

Conclusions: State-of-the-art NLP models can accurately extract CKD-related features from ultrasound reports, with the potential of providing a scalable tool for early detection and risk stratification. Future research should focus on validating these models across different healthcare systems.

Background: Acute kidney injury (AKI) usually occurs after cardiopulmonary bypass (CPB) and threatens life without timely intervention. Early assessment and prevention are critical for saving AKI patients. However, numerical data-driven models make it difficult to predict the AKI risk using preoperative data and lack preventive measures. Large language models (LLM) have demonstrated significant potential for medical decision-making, offering a promising approach.

Objective: For preoperative assessment and prevention of CPB-associated AKI (CPB-AKI).

Methods: Clinical variables were converted into text through prompt engineering and a LLM was used to extract information hidden in the semantics of subtle changes. A multimodal fusion model, fuzing semantic and numerical information, was proposed to assess the AKI risk before surgery. We then used a structural equation model to analyze the impact of preoperative features and intraoperative interventions on CPB-AKI prevention.

Results: A total of 2,056 patients who underwent CPB were enrolled from the intensive care unit of Sir Run Run Shaw Hospital between 2014 and 2022, with 40.5% progressing to AKI. Our model performed better with an area under the receiver operating characteristic curve of 0.9201 compared with the baseline models. The structural equation model's chi-square to degrees of freedom ratio was 0.46, less than 2.0. We discussed how the preoperative prediction model could optimize intraoperative interventions to prevent CPB-AKI.

Conclusions: The prediction model can predict CPB-AKI risk earlier after fuzing the clinical characteristics and their semantics. Preoperative assessment and intraoperative interventions provide decision-making to prevent CPB-AKI.

Background: To explore the associations of age, intramuscular adipose tissue index (IATI), and serum albumin with survival status in initial dialysis patients and the mediating effects.

Methods: Totally 1,044 Chinese initial dialysis patients from four hospitals (2014-2020) were eventually enrolled and followed up to December 31, 2022 or until death in this retrospective cohort study. IATI was defined as the ratio of low attenuation muscle density to skeletal muscle density assessed by CT at the first lumbar vertebra level. Multivariate Cox regression and two-piecewise Cox proportional hazards models were used to determine the risk factors for all-cause mortality and to perform stratified analysis. Mediation analysis was conducted to identify mediators.

Results: High IATI, age > 60 years, and low serum albumin were significant independent risk factors for all-cause mortality. The association between IATI and all-cause mortality remained significant in female patients, and those with low neutrophil/lymphocyte ratios, or without coronary heart disease. When age and IATI were categorical variables, age had a significant indirect effect on all-cause mortality (0.015) and survival time (-1.262) via IATI, while IATI indirectly influenced all-cause mortality through serum albumin (0.012).

Conclusions: Age > 60 years and high IATI are risk factors for all-cause mortality while serum albumin is protective in initial dialysis patients. The relationship between age and survival status may be mediated by IATI, while the effect of IATI on all-cause mortality may be mediated by serum albumin.

Large language models (LLMs) represent a transformative advance in artificial intelligence, with growing potential to impact chronic kidney disease (CKD) management. CKD is a complex, highly prevalent condition requiring multifaceted care and substantial patient engagement. Recent developments in LLMs-including conversational AI, multimodal integration, and autonomous agents-offer novel opportunities to enhance patient education, streamline clinical documentation, and support decision-making across nephrology practice. Early reports suggest that LLMs can improve health literacy, facilitate adherence to complex treatment regimens, and reduce administrative burdens for clinicians. However, the rapid deployment of these technologies raises important challenges, including patient privacy, data security, model accuracy, algorithmic bias, and ethical accountability. Moreover, real-world evidence supporting the safety and effectiveness of LLMs in nephrology remains limited. Addressing these challenges will require rigorous validation, robust regulatory frameworks, and ongoing collaboration between clinicians, AI developers, and patients. As LLMs continue to evolve, future efforts should focus on the development of nephrology-specific models, prospective clinical trials, and strategies to ensure equitable and transparent implementation. If appropriately integrated, LLMs have the potential to reshape the landscape of CKD care and education, improving outcomes for patients and supporting the nephrology workforce in an era of increasing complexity.

This study aimed to identify potential serum biomarkers and explore associated signaling pathways involved in cardiac surgery-associated acute kidney injury (CSA-AKI) by integrating proteomic and lipidomic analyses. A total of 34 patients were enrolled, including 17 CSA-AKI patients and 17 controls. Untargeted lipidomic analysis was performed using liquid chromatography-mass spectrometry (LC-MS) approach. Proteomics analysis was conducted using data-independent acquisition-based LC-MS/MS detection. The integration of proteomics and lipidomics was evaluated using statistical and bioinformatics methods. The two groups had different serum protein and lipid profiles, which included 185 differentially expressed proteins and 65 differentially expressed lipids. The protein and lipid molecules were enriched in biologic pathway implicated in biosynthesis of arachidonic acid metabolism, gonadotropin-releasing hormone signaling pathway, leukocyte trans endothelial migration, and glycerophospholipid metabolism. Correlation analysis revealed that 4 proteins were positively correlated with 21 lipids, whereas 7 proteins were negatively correlated with the 21 lipids. These results suggested that significantly altered proteins and lipids may be involved in the early stage of CSA-AKI and could serve as potentially promising markers. The association between proteins and lipid molecules and the underlying signaling pathways may elucidate the pathogenesis of CSA-AKI.

Ensuring fluent extracorporeal circulation and preventing circuit clotting are important for end-stage kidney disease (ESKD) patients undergoing continuous renal replacement therapy (CRRT). This study aimed to develop a predictive model using machine learning (ML) algorithms to evaluate clotting risk after initiating CRRT, enhancing treatment safety and effectiveness. This study involved 636 ESKD patients who underwent CRRT. Feature selection was conducted via the least absolute shrinkage and selection operator (LASSO) algorithm. ML algorithms, including support vector machine (SVM), extreme gradient boosting (XGBoost), random forest (RF), gradient boosting machine (GBM), decision tree, and logistic regression (LR), were applied to construct models through tenfold cross-validation. Model performance was assessed via the area under the receiver operating characteristic curve (AUC) and additional metrics. The Shapley additive explanation (SHAP) values quantify each feature's contribution. This study included 199 patients with blood clots during extracorporeal circulation, corresponding to an incidence rate of 31.3%. The AUC values were 0.864 (SVM), 0.815 (XGBoost), 0.806 (GBM), 0.778 (RF), 0.732 (Decision Tree), and 0.717 (LR). The SVM exhibited the best performance. The initial dose of low-molecular-weight heparin (LMWH) was identified as the most significant factor influencing coagulation. ML serves as a reliable tool for predicting the risk of extracorporeal circuit clotting in ESKD patients undergoing CRRT. The SHAP method elucidates key risk factors, providing a basis for early clinical intervention.

Background and aims: Lipid accumulation in podocytes is a major driver of diabetic kidney disease (DKD). Hypoxia-inducible factor 2α (HIF-2α) plays an important role in regulating metabolism. The function of HIF-2α in lipid metabolism in podocytes and the progression of DKD remain unclear.

Methods: We investigated the effects of HIF-2α on podocyte damage and lipid metabolism using immunofluorescence, flow cytometry, ELISA, and Western blotting. In order to characterize the regulatory effects of HIF-2α, we also used ChIP and dual-luciferase reporter assays to investigate the role of sphingosine kinase 1 (SPHK1), a crucial enzyme in sphingosine-1-phosphate (S1P) synthesis. In vivo, the effect of HIF-2α on lipid metabolism disorders in db/db mice was investigated using the HIF-2α inhibitor PT-2385.

Results: Our results revealed that HIF-2α overexpression improved lipid metabolism in DKD by enhancing cholesterol efflux via reduced S1P synthesis in podocytes by 25.69%. Inhibition of HIF-2α expression in the mouse model of diabetes exacerbated podocyte damage and proteinuria. Inhibition of SPHK1 expression rescued HIF-2α knockdown-mediated lipid disorders in podocytes. HIF-2α inhibited the transcription of SPHK1 by binding to the promoter region of SPHK1 and reduced S1P synthesis. Furthermore, we found that FG-4592, a HIF prolyl hydroxylase inhibitor, reduced the total cholesterol level in DKD by activating HIF-2α, thereby protecting against DKD.

Conclusion: HIF-2α ameliorated lipid metabolism disorders and podocyte damage in DKD by downregulating S1P, providing a novel insight for HIF-2α against DKD.

Background: Mitochondrial dysfunction is linked to hyperuricemia (HUA), but its genetic pathophysiology is not yet fully understood. We employed Mendelian randomization (MR) to integrate multi-omics data and explore the associations between mitochondrial-related genes and HUA.

Methods: We conducted a summary data-based MR analysis to investigate potential targets associated with HUA by integrating mitochondrial-related DNA methylation, gene expression, and protein quantitative trait loci. Additionally, to further explore the potential associations between DNA methylation, gene expression, and protein abundance, we performed MR and co-localization analyses to examine causal relationships between candidate gene methylation and expression, as well as between gene expression and protein abundance.

Result: Through the integration of multi-omics evidence, we identified one primary gene, NUDT2, and three secondary genes, BOLA1, COMT, and HAGH. At the protein level, NUDT2 and COMT are negatively correlated with HUA risk, whereas BOLA1 and HAGH are positively correlated with HUA risk. Our analysis revealed a positive correlation between the methylation of cg06605933 in BOLA1 and its protein levels, which aligns with the negative effect of cg06605933 methylation on HUA risk. Additionally, we observed a positive correlation between NUDT2 gene expression and protein levels, confirming its beneficial effect on HUA risk. Strong co-localization support was found between the methylation of cg06605933 (PPH4 = 85.1%) in BOLA1 and protein abundance, as well as between NUDT2 gene expression (PPH4 = 96.6%) and protein levels.

Conclusion: Our study identified mitochondrial genes NUDT2, BOLA1, COMT, and HAGH as potentially associated with HUA risk, supported by evidence from various omics levels.

Background: Perioperative hyperamylasemia has been observed in several kidney transplant recipients (KTRs) at our center. However, there are currently no published reports on this observation. This study aimed to identify the risk factors associated with perioperative hyperamylasemia in KTRs.

Methods: The data from 540 deceased-donor kidney recipients in our hospital from January 2020 to December 2023 were retrospectively analyzed. Variables such as gender, past medical history, relevant laboratory tests, and trough concentration of calcineurin inhibitors (CNIs) at the time of serum amylase maximum were collected for all patients. Univariate and multivariate logistic regression analyses were used to determine the risk factors associated with perioperative hyperamylasemia.

Results: Among all KTRs, 153 patients (28.3%) developed perioperative hyperamylasemia. Multivariate logistic regression analysis indicated that preoperative serum phosphate (odds ratio [OR] = 1.62, 95% confidence interval [CI]: 1.160 - 2.266, p = 0.005), preoperative serum amylase (OR = 1.01, 95% CI: 1.006 - 1.015, p < 0.001), and high perioperative CNIs trough concentration (OR = 2.335, 95% CI: 1.560 - 3.494, p < 0.001) were risk factors associated with perioperative hyperamylasemia. In addition, we used all three as a hybrid model to predict perioperative hyperamylasemia, which demonstrated good predictive value (area under the ROC curve [AUC] = 0.687, 95% CI: 0.64-0.734).

Conclusion: Elevated preoperative serum phosphate levels, preoperative serum amylase levels, and high perioperative CNIs trough concentrations are risk factors for perioperative hyperamylasemia. This study may provide valuable insights for clinicians to identify the causes of perioperative hyperamylasemia and formulate prevention and treatment strategies.

Objective: To compare the efficacy and safety of ticagrelor versus clopidogrel in stroke patients who were CYP2C19 loss-of-function (LOF) carriers stratified by age and renal function.

Methods: Patients in the CHANCE-2 trial were randomized to ticagrelor-aspirin or clopidogrel-aspirin treatment. The primary efficacy outcome was occurrence of a new stroke within 90 days, while bleeding was assessed for safety. Patients were categorized based on age and estimated glomerular filtration rate (eGFR).

Results: In patients with eGFR >90 mL/min/1.73 m², ticagrelor-aspirin was associated with a significantly lower risk of the subsequent stroke within 90 days compared with the clopidogrel-aspirin in those aged over 65 years (HR 0.53, 95% CI 0.33-0.85, p = 0.008) and under 65 years (HR, 0.67, 95% CI, 0.47-0.96, p = 0.03). While in those with eGFR 60-89 mL/min/1.73 m², ticagrelor did not show superiority over clopidogrel in reducing stroke regardless of age category (age ≥ 65: HR 1.14, 95% CI 0.71-1.84, p = 0.59; age < 65: HR 0.40, 95% CI 0.12-1.33, p = 0.13). The incidence of mild bleeding events was higher with ticagrelor-aspirin treatment in those aged < 65 years with eGFR ≥90 mL/min/1.73 m² (HR 3.33, 95% CI 2.18-5.10, p < 0.001) and in those aged ≥ 65 years with eGFR <60mL/min/1.73 m² (HR 8.68, 95% CI 1.06-71.1, p = 0.04).

Conclusions: Elderly patients with normal renal function appear to benefit from ticagrelor compared with clopidogrel. Both younger patients with normal renal function and those with advanced age and renal insufficiency are prone to mild bleeding.