Min Yang

Background: Interleukin-21 (IL21) is an important immunoregulatory cytokine involved in antitumor immune response; however, its cellular source, regulatory program, and functional significance in colorectal cancer remain incompletely understood. Methods: We analyzed tumor-infiltrating T cells from the public single-cell RNA sequencing dataset to characterize IL21 expression patterns in colorectal cancer. External validation was performed using a single-cell cohort of colorectal cancer cells treated with anti-PD-1 neoadjuvant therapy. Multiplex immunofluorescence staining was performed on colorectal cancer tissue microarrays to validate the spatial distribution and cellular origin of IL21. Correlation analysis was used to assess the relationship between IL21 and immune-related cell subsets. SCENIC analysis was performed to identify candidate transcription factors associated with CXCL13-associated CD4 ⁺ T cell subsets. Furthermore, tumor tissues from an AOM/DSS-induced mouse colorectal cancer model were cultured in vitro (with or without IL21) and batch RNA sequenced to explore IL21-induced transcriptional changes. Results: Single-cell analysis showed that IL21 expression was restricted to a limited fraction of tumor-infiltrating T cells and was predominantly enriched in CD4 ⁺ T cell subsets, particularly CXCL13 ⁺ CD4 ⁺ T cell populations, whereas IL21R was more broadly expressed. Analysis of the PD-1 anti-neoadjuvant therapy cohort further confirmed that IL21 is also preferentially expressed in specialized CD4 ⁺ T cell subsets, including CXCL13-associated cell populations. Tissue-level multiplex immunofluorescence confirmed that IL21, IL21 ⁺ CD4 ⁺ T cells, and CXCL13 ⁺ IL21 ⁺ CD4 ⁺ T cells were all significantly increased in tumor tissues compared with adjacent normal tissues. Correlation analysis further suggested that IL21 expression was positively associated with CD4 ⁺ , CXCL13 ⁺ , IL21 ⁺ CD4 ⁺ , and CXCL13 ⁺ IL21 ⁺ CD4 ⁺ T cell populations. SCENIC analysis identified a distinct regulon landscape in the CD4_C09-CXCL13 subset, with STAT1, IRF9, IRF7, TBX21, and IRF4 emerging as candidate key transcription factors. Functionally, exogenous IL21 induced marked transcriptional remodeling in murine colorectal tumor tissues, characterized by activation of immune- and cytokine-related pathways and suppression of multiple Wnt-associated genes. Conclusions: Our study identifies CXCL13-associated CD4 ⁺ T cells as a major IL21-producing population in colorectal cancer and links this axis to an immune-active tumor microenvironment. Although direct mechanistic validation remains necessary, our findings provide a cellular and transcriptional framework for future investigation of IL21-mediated immune regulation in colorectal cancer.

Background: Peritoneal dialysis-associated peritonitis (PDAP) remains a major complication of peritoneal dialysis (PD). The controlling nutritional status (CONUT) score, which reflects the immune-nutritional state, may offer predictive value in identifying patients at risk. This study aimed to evaluate the utility of machine learning models in predicting early-onset PDAP and to assess the prognostic importance of baseline CONUT score, 6-month CONUT score, and their dynamic changes. Methods: In this multicenter prospective cohort study, 675 patients initiating PD were enrolled. Multivariable logistic regression was performed to identify clinical predictors of early-onset peritonitis, while Kaplan–Meier survival analysis was used to compare peritonitis-free survival among patients with no peritonitis, early-onset peritonitis, and late-onset peritonitis. To enhance predictive performance, machine learning models including XGBoost, LightGBM, and their ensemble were constructed. Feature selection was based on SHapley Additive exPlanations (SHAP) values derived from an initial XGBoost model. The top 10 SHAP-ranked features were used to train all models. Model performance was assessed using area under the receiver operating characteristic curve (AUC), and SHAP summary plots were generated to interpret feature contributions. Results: Over a median follow-up period of 41.8 months, 82 patients developed early-onset PDAP. Multivariable logistic regression identified baseline total cholesterol, neutrophil-to-lymphocyte ratio, and 6-month CONUT score as independent predictors of early-onset PDAP (vs. no PDAP; p < 0.05). In comparisons between early- and late-onset PDAP, older age, longer PD duration, and lower 6-month CONUT score were independently associated with a decreased likelihood of early-onset PDAP ( p < 0.05). Using the top 10 SHAP-ranked features, three models (XGBoost, LightGBM, and an ensemble) were trained. For distinguishing early-onset PDAP from no PDAP, LightGBM performed best (AUC = 0.717), followed by the ensemble (0.698) and XGBoost (0.670). In differentiating early- from late-onset PDAP, LightGBM showed the highest AUC (0.781), outperforming the ensemble (0.744) and XGBoost (0.691). SHAP summary plots consistently identified the 6-month CONUT score as the important feature across both classification tasks. Conclusion: The 6-month CONUT score is an independent predictor of early-onset PDAP and was among the top contributing features in multiple machine learning models. Integrating SHAP-based feature selection with gradient boosting improved model accuracy and interpretability. Dynamic monitoring of nutritional-immune status may aid in early risk stratification and guide personalized prevention strategies in patients undergoing PD.

Purpose: Diabetic kidney disease (DKD) is a major contributor to chronic kidney disease worldwide. Bile acids (BAs) are increasingly recognized as key regulators of glucose metabolism and kidney function. This study aimed to investigate the role of BA metabolism in the progression of DKD. Methods: Plasma BA profiles were measured in healthy controls (HC), patients with type 2 diabetes mellitus (T2DM), and patients with DKD using ultra-high-performance liquid chromatography–tandem mass spectrometry (UPLC-MS/MS). After identifying potential BA biomarkers in the clinical cohort, in vivo validation was conducted using dehydrolithocholic acid (DHLCA) intervention in DKD mouse model. Kidney injury markers, as well as the expression of Takeda G protein-coupled receptor 5 (TGR5) and farnesoid X receptor (FXR), were evaluated. In addition, gut microbiota (GM) composition was analyzed via metagenomic sequencing following DHLCA treatment. Results: The plasma DHLCA levels were significantly lower in DKD with macroalbuminuria group compared to T2DM group and DKD with microalbuminuria group (P < 0.01). Partial Spearman correlation analysis adjusted for age and diabetes duration showed that DHLCA levels were negatively correlated with urine albumin (ρ = – 0.347; 95% CI, – 0.531 to – 0.135; q = 0.008) and urine albumin-to-creatinine ratio (UACR) (ρ = – 0.332; 95% CI, – 0.499 to – 0.155; q = 0.010). In vivo, DHLCA administration significantly reduced UACR and fasting blood glucose (FBG) levels (P < 0.01), and improved liver function (ALT, P < 0.05) in DKD mice. DHLCA treatment attenuated renal tubular injury, restored TGR5 and FXR expression in kidney tissue, and decreased levels of kidney injury molecule-1 (KIM-1) and neutrophil gelatinase-associated lipocalin (NGAL). Metagenomic analysis revealed an enrichment of Lachnospiraceae bacterium following DHLCA treatment. Conclusion: DHLCA may represent a promising therapeutic candidate for DKD by targeting the TGR5/FXR signaling pathway and GM remodeling. Its metabolic and kidney benefits, along with an improved hepatic profile and absence of hepatotoxicity, support further translational investigation.

Aim: The aim of this study was to develop a predictive model for the progression of diabetic kidney disease (DKD) to end-stage renal disease (ESRD) and to evaluate the effectiveness of renal pathology and the kidney failure risk equation (KFRE) in this context. Methods: The study comprised two parts. The first part involved 555 patients with clinically diagnosed DKD, while the second part focused on 85 patients with biopsy-proven DKD. Cox regression analysis and competing risk regression were employed to identify independent predictors. Time-dependent receiver operating characteristic (ROC) was used to evaluate prediction performance, and the area under the curve (AUC) was calculated to assess the model’s accuracy. Results: The Cox regression model developed for the 555 patients clinically diagnosed with DKD identified 5 predictors (body mass index (BMI), estimated glomerular filtration rate (eGFR), 24-hour urinary total protein (UTP), systemic immune-inflammatory index (SII), and controlling nutritional status (CONUT), whereas the Competing risks model included 4 predictors (BMI, eGFR, UTP, CONUT). Among 85 patients with biopsy-proven diabetic DKD, the combined prognostic model integrating KFRE, interstitial fibrosis and tubular atrophy (IFTA), SII and BMI demonstrated enhanced predictive ability at 5 years. The developed models offer improved accuracy over existing methods by incorporating renal pathology and novel inflammatory indices, making them more applicable in clinical settings. Conclusion: The predictive model proved to be effective in assessing the progression of DKD to ESRD. Additionally, the combined model of KFRE, IFTA, SII, and BMI demonstrates high predictive performance. Future studies should validate these models in larger cohorts and explore their integration into routine clinical practice to enhance personalized risk assessment and management.

Although the patient survival rate for many malignancies has been improved with immune checkpoint inhibitors (ICIs), some patients experience various immune-related adverse events (irAEs). IrAEs impact several organ systems, including the kidney. With anti-programmed cell death protein 1 (PD-1) therapy (pembrolizumab), kidney-related adverse events occur relatively rarely compared with other irAEs. However, the occurrence of AKI usually leads to anti-PD-1 therapy interruption or discontinuation. Therefore, there is an urgent need to clarify the mechanisms of renal irAEs (R-irAEs) to facilitate early management. This study aimed to analyse the characteristics of peripheral blood mononuclear cells (PBMCs) in R-irAEs. PBMCs were collected from three patients who developed R-irAEs after anti-PD-1 therapy and three patients who did not. The PBMCs were subjected to scRNA-seq to identify cell clusters and differentially expressed genes (DEGs). Kyoto Encyclopedia of Genes and Genomes (KEGG) and gene ontology (GO) enrichment analyses were performed to investigate the most active biological processes in immune cells. Fifteen cell clusters were identified across the two groups. FOS, RPS26, and JUN were the top three upregulated genes in CD4+ T cells. The DEGs in CD4+ T cells were enriched in Th17 differentiation, Th1 and Th2 cell differentiation, NF-kappa B, Nod-like receptor, TNF, IL-17, apoptosis, and NK cell-mediated cytotoxicity signaling pathways. RPS26, TRBV25-1, and JUN were the top three upregulated genes in CD8+ T cells. The DEGs in CD8+ T cells were enriched in Th17 cell differentiation, antigen processing and presentation, natural killer cell-mediated cytotoxicity, the intestinal immune network for IgA production, the T-cell receptor signalling pathway, Th1 and Th2 cell differentiation, the phagosome, and cell adhesion molecules. In conclusion, R-irAEs are associated with immune cell dysfunction. DEGs and their enriched pathways identified in CD4+ T cells and CD8+ T cells play important roles in the development of renal irAEs related to anti-PD-1 therapy. These findings offer fresh perspectives on the pathogenesis of renal damage caused by anti-PD-1 therapy.

Purpose This retrospective study aimed to investigate the relationship between alternative insulin resistance (IR) indexes not reliant on insulin and diabetic kidney disease (DKD) incidence in a newly diagnosed cohort of individuals with type 2 diabetes mellitus (T2DM). Methods We conducted a retrospective analysis of baseline characteristics in a cohort of 521 individuals with T2DM, then followed up on the outcome of DKD. To assess the predictive ability of IR indexes, we compared the performance of four non-insulin-based IR indexes and the homeostasis model for insulin resistance (HOMA-IR) using logistic regression and consistency-statistics (C-statistics). Furthermore, we computed the net reclassification index (NRI) and integrated discrimination improvement (IDI) to evaluate the additional effects of the indexes. Results The four alternative IR indexes of DKD patients were significantly higher than those of non-DKD. After adjustment for other variables, the highest tertile of all indexes was significantly related to DKD incidence, compared with the lowest tertile. Furthermore, the C-statistics for the triglyceride-glucose index (TyG index) and triglyceride to high-density lipoprotein ratio (TG/HDL) were all 0.652, while triglyceride glucose-body mass index (TyG-BMI) and metabolic score for insulin resistance (METS-IR) were 0.639 and 0.651, respectively. The incorporation of the alternative IR indexes into the baseline model revealed positive additional effects, leading to an improved prediction of the risk for DKD. Conclusions It was discovered that the alternative IR indexes served as independent risk factors of DKD. Among the four alternative indexes, TyG index and TG/HDL had the best prediction performance for DKD, followed by METS-IR.

Background: Diabetic kidney disease (DKD), a common microvascular complication of diabetes mellitus (DM), is always asymptomatic until it develops to the advanced stage. Thus, we aim to develop a nomogram prediction model for progression to DKD in newly diagnosed type 2 diabetes mellitus (T2DM). Methods: This was a single-center analysis of prospective data collected from 521 newly diagnosed patients with T2DM. All related clinical records were incorporated, including the triglyceride-glucose index (TyG index). The least absolute shrinkage and selection operator (LASSO) was used to build a prediction model. In addition, discrimination, calibration, and clinical practicality of the nomogram were evaluated. Results: In this study, 156 participants were incorporated as the validation set, while the remaining 365 were incorporated into the training set. The predictive factors included in the individualized nomogram prediction model included 5 variables. The area under the curve (AUC) for the prediction model was 0.826 (95% CI 0.775 to 0.876), indicating excellent discrimination performance. The model performed exceptionally well in terms of predictive accuracy and clinical applicability, according to calibration curves and decision curve analysis. Conclusion: The predictive nomogram for the risk of DKD in newly diagnosed T2DM patients had outstanding discrimination and calibration, which could help in clinical practice.

Background: The risk of cardiovascular disease (CVD) in diabetes mellitus (DM) patients is two- to three-fold higher than in the general population. We designed a 10-year cohort trial in T2DM patients to explore the performance of QRESEARCH risk estimator version 3 (QRISK3) as a CVD risk assessment tool and compared to Framingham Risk Score (FRS).Method: This is a single-center analysis of prospective data collected from 566 newly-diagnosed patients with type 2 DM (T2DM). The risk scores were compared to CVD development in patients with and without CVD. The risk variables of CVD were identified using univariate analysis and multivariate cox regression analysis. The number of patients classified as low risk (20%) for two tools were identified and compared, as well as their sensitivity, specificity, positive and negative predictive values, and consistency (C) statistics analysis.Results: Among the 566 individuals identified in our cohort, there were 138 (24.4%) CVD episodes. QRISK3 classified most CVD patients as high risk, with 91 (65.9%) patients. QRISK3 had a high sensitivity of 91.3% on a 10% cut-off dichotomy, but a higher specificity of 90.7% on a 20% cut-off dichotomy. With a 10% cut-off dichotomy, FRS had a higher specificity of 89.1%, but a higher sensitivity of 80.1% on a 20% cut-off dichotomy. Regardless of the cut-off dichotomy approach, the C-statistics of QRISK3 were higher than those of FRS.Conclusion: QRISK3 comprehensively and accurately predicted the risk of CVD events in T2DM patients, superior to FRS. In the future, we need to conduct a large-scale T2DM cohort study to verify further the ability of QRISK3 to predict CVD events.