Predicting diabetic kidney disease for type 2 diabetes mellitus by machine learning in the real world: a multicenter retrospective study DOI Creative Commons

X Liu,

Minjie Duan,

Hao Huang

et al.

Frontiers in Endocrinology, Journal Year: 2023, Volume and Issue: 14

Published: July 4, 2023

Objective Diabetic kidney disease (DKD) has been reported as a main microvascular complication of diabetes mellitus. Although renal biopsy is capable distinguishing DKD from Non disease(NDKD), no gold standard validated to assess the development DKD.This study aimed build an auxiliary diagnosis model for type 2 (T2DKD) based on machine learning algorithms. Methods Clinical data 3624 individuals with (T2DM) was gathered January 1, 2019 December 31, using multi-center retrospective database. The fell into training set and validation at random ratio 8:2. To identify critical clinical variables, absolute shrinkage selection operator lowest number employed. Fifteen models were built support T2DKD, optimal selected in accordance area under receiver operating characteristic curve (AUC) accuracy. improved use Bayesian Optimization methods. Shapley Additive explanations (SHAP) approach used illustrate prediction findings. Results diagnosed 1856 (51.2 percent) within final cohort. As revealed by SHAP findings, Categorical Boosting (CatBoost) achieved performance 1in risk AUC 0.86 top 38 characteristics. findings suggested that simplified CatBoost 0.84 12 more basic features consisted systolic blood pressure (SBP), creatinine (CREA), length stay (LOS), thrombin time (TT), Age, prothrombin (PT), platelet large cell (P-LCR), albumin (ALB), glucose (GLU), fibrinogen (FIB-C), red distribution width-standard deviation (RDW-SD), well hemoglobin A1C(HbA1C). Conclusion A learning-based developing T2DKD built, its effectiveness verified. can contribute T2DKD. Clinicians could gain insights outcomes if ML made interpretable.

Language: Английский

The single-cell transcriptomic landscape of early human diabetic nephropathy DOI Creative Commons
Parker C. Wilson, Hao Wu, Yuhei Kirita

et al.

Proceedings of the National Academy of Sciences, Journal Year: 2019, Volume and Issue: 116(39), P. 19619 - 19625

Published: Sept. 10, 2019

Diabetic nephropathy is characterized by damage to both the glomerulus and tubulointerstitium, but relatively little known about accompanying cell-specific changes in gene expression. We performed unbiased single-nucleus RNA sequencing (snRNA-seq) on cryopreserved human diabetic kidney samples generate 23,980 transcriptomes from 3 control early samples. All major cell types of were represented final dataset. Side-by-side comparison demonstrated cell-type-specific expression that are important for ion transport, angiogenesis, immune activation. In particular, we show thick ascending limb, late distal convoluted tubule, principal cells all adopt a signature consistent with increased potassium secretion, including alterations Na+/K+-ATPase, WNK1, mineralocorticoid receptor, NEDD4L expression, as well decreased paracellular calcium magnesium reabsorption. also identify strong angiogenic signatures glomerular types, proximal cells. Taken together, these results suggest secretion signaling represent responses nephropathy.

Language: Английский

Citations

422

Diabetic vascular diseases: molecular mechanisms and therapeutic strategies DOI Creative Commons

Yiwen Li,

Yanfei Liu, Shiwei Liu

et al.

Signal Transduction and Targeted Therapy, Journal Year: 2023, Volume and Issue: 8(1)

Published: April 10, 2023

Abstract Vascular complications of diabetes pose a severe threat to human health. Prevention and treatment protocols based on single vascular complication are no longer suitable for the long-term management patients with diabetes. Diabetic panvascular disease (DPD) is clinical syndrome in which vessels various sizes, including macrovessels microvessels cardiac, cerebral, renal, ophthalmic, peripheral systems diabetes, develop atherosclerosis as common pathology. Pathological manifestations DPDs usually manifest macrovascular atherosclerosis, well microvascular endothelial function impairment, basement membrane thickening, microthrombosis. Cardiac, microangiopathy coexist microangiopathy, while renal retinal predominantly microangiopathic. The following associations exist between DPDs: numerous similar molecular mechanisms, risk-predictive relationships diseases. Aggressive glycemic control combined early comprehensive intervention key prevention treatment. In addition widely recommended metformin, glucagon-like peptide-1 agonist, sodium-glucose cotransporter-2 inhibitors, latest aldose reductase peroxisome proliferator-activated receptor-γ agonizts, glucokinases mitochondrial energy modulators, etc. under active development. proposed obtain more systematic care requires center focusing This would leverage advantages cross-disciplinary approach achieve better integration pathogenesis therapeutic evidence. Such strategy confer benefits promote development DPD discipline.

Language: Английский

Citations

257

Genome-Wide Association Study of Diabetic Kidney Disease Highlights Biology Involved in Glomerular Basement Membrane Collagen DOI Open Access
Rany M. Salem, Jennifer N. Todd, Niina Sandholm

et al.

Journal of the American Society of Nephrology, Journal Year: 2019, Volume and Issue: 30(10), P. 2000 - 2016

Published: Sept. 19, 2019

Significance Statement Although studies show that diabetic kidney disease has a heritable component, searches for the genetic determinants of this complication diabetes have had limited success. In study, new international genomics consortium, JDRF funded Diabetic Nephropathy Collaborative Research Initiative, assembled nearly 20,000 samples from participants with type 1 diabetes, and without disease. The authors found 16 disease–associated loci at genome-wide significance. strongest signal centers on protective missense coding variant COL4A3 , gene encodes component glomerular basement membrane that, when mutated, causes progressive inherited nephropathy Alport syndrome. These GWAS-identified risk may provide insights into pathogenesis help identify potential biologic targets prevention treatment. Background demonstrates both familial clustering single nucleotide polymorphism heritability, specific factors influencing remain largely unknown. Methods To variants predisposing to disease, we performed association study (GWAS) analyses. Through collaboration Diabetes large collection cohorts harmonized phenotypes. We used spectrum ten definitions based albuminuria renal function. Results Our GWAS meta-analysis included results up 19,406 individuals European descent diabetes. identified significant loci. (rs55703767) is common mutation in collagen IV alpha 3 chain ( COL4A3) gene, which major structural (GBM). Mutations are implicated nephropathies, including rs55703767 minor allele (Asp326Tyr) against several ESKD, demonstrated GBM width; carriers thinner before any signs its effect was dependent glycemia. Three other or near genes known suggestive involvement condition BMP7) biology COLEC11 DDR1 ). Conclusions novel

Language: Английский

Citations

171

Pathophysiologic Mechanisms and Potential Biomarkers in Diabetic Kidney Disease DOI Creative Commons
Chan‐Young Jung, Tae‐Hyun Yoo

Diabetes & Metabolism Journal, Journal Year: 2022, Volume and Issue: 46(2), P. 181 - 197

Published: March 25, 2022

Although diabetic kidney disease (DKD) remains the leading cause of end-stage eventually requiring chronic replacement therapy, prevalence DKD has failed to decline over past 30 years. In order reduce prevalence, extensive research been ongoing improve prediction onset and progression. most commonly used markers are albuminuria estimated glomerular filtration rate, their limitations have encouraged researchers search for novel biomarkers that could risk stratification. Considering is a complex process involves several pathophysiologic mechanisms such as hyperglycemia induced inflammation, oxidative stress, tubular damage, damage fibrosis, many capture one specific mechanism developed. Moreover, increasing use high-throughput omic approaches analyze biological samples include proteomics, metabolomics, transcriptomics emerged strong tool in biomarker discovery. This review will first describe recent advances understanding pathophysiology DKD, second, current clinical well status multiple potential with respect protein biomarkers, transcriptomics.

Language: Английский

Citations

118

Association between neutrophil-to-lymphocyte ratio and diabetic kidney disease in type 2 diabetes mellitus patients: a cross-sectional study DOI Creative Commons
Xiaowan Li, Lanyu Wang, Min Liu

et al.

Frontiers in Endocrinology, Journal Year: 2024, Volume and Issue: 14

Published: Jan. 4, 2024

Aims This investigation examined the possibility of a relationship between neutrophil-to-lymphocyte ratio (NLR) and diabetic kidney disease (DKD) in type 2 diabetes mellitus (T2DM) patients. Methods Adults with T2DM who were included National Health Nutrition Examination Survey (NHANES) 1999 2020 subjects current cross-sectional investigation. Low estimated glomerular filtration rate (eGFR) (< 60 mL/min/1.73 m ) or albuminuria (urinary albumin-to-creatinine (ACR) ≥ 30 mg/g) patients diagnostic criteria for DKD. Weighted multivariable logistic regression models generalized additive used to investigate independent relationships NLR levels DKD, albuminuria, low-eGFR. Additionally, we low-eGFR other inflammatory markers, such as aggregate index systemic inflammation (AISI), immune-inflammation (SII), system response (SIRI), platelet-to-lymphocyte (PLR) monocyte-to-lymphocyte (MLR). Their capabilities evaluated contrasted using receiver operating characteristic (ROC) curves. Results 44.65% 7,153 participants recruited this study males. prevalent 31.76%, 23.08%, 14.55% cases, respectively. Positive correlations seen prevalences Subgroup analysis interaction tests revealed that associations not significantly different across populations. In addition, MLR, SII SIRI showed positive prevalence ROC discovered when compared markers (MLR, PLR, SII, SIRI, AISI), may demonstrate more discriminatory power accuracy assessing risk Conclusion Compared serve effective potential marker identifying US elevated NLR, should be closely monitored their renal function.

Language: Английский

Citations

30

Association of Multiple Plasma Biomarker Concentrations with Progression of Prevalent Diabetic Kidney Disease: Findings from the Chronic Renal Insufficiency Cohort (CRIC) Study DOI Open Access
Sarah J. Schrauben, Haochang Shou, Xiaoming Zhang

et al.

Journal of the American Society of Nephrology, Journal Year: 2020, Volume and Issue: 32(1), P. 115 - 126

Published: Oct. 29, 2020

Although diabetic kidney disease is the leading cause of ESKD in United States, identifying those patients who progress to difficult. Efforts are under way determine if plasma biomarkers can help identify these high-risk individuals.In our case-cohort study 894 Chronic Renal Insufficiency Cohort Study participants with diabetes and an eGFR <60 ml/min per 1.73 m2 at baseline, were randomly selected for subcohort; cases developed progressive (ESKD or 40% decline). Using a multiplex system, we assayed related tubular injury, inflammation, fibrosis (KIM-1, TNFR-1, TNFR-2, MCP-1, suPAR, YKL-40). Weighted Cox regression models progression disease, mixed-effects estimated biomarker relationships rate change.Median follow-up was 8.7 years. Higher concentrations KIM-1, YKL-40 each associated greater risk even after adjustment established clinical factors. After accounting competing biomarkers, remained disease; TNFR-2 had highest (adjusted hazard ratio, 1.61; 95% CI, 1.15 2.26). decline.Higher levels increased other biomarkers. These findings validate previous literature on KIM-1 prevalent CKD provide new insights into influence suPAR as that require validation.

Language: Английский

Citations

110

Effects of Selonsertib in Patients with Diabetic Kidney Disease DOI Open Access
Glenn M. Chertow, Pablo E. Pérgola, Chen Fang

et al.

Journal of the American Society of Nephrology, Journal Year: 2019, Volume and Issue: 30(10), P. 1980 - 1990

Published: Sept. 10, 2019

Significance Statement Findings in animal models of diabetic kidney disease identified selonsertib, a selective inhibitor apoptosis signal–regulating kinase 1 (ASK1), as potential therapeutic agent. In randomized, dose-ranging, placebo-controlled phase 2 trial evaluating selonsertib’s safety and efficacy patients with moderate-to-advanced disease, the authors found that selonsertib appeared safe, no dose-dependent adverse effects over 48 weeks, including for 18-mg daily dose thought to maximally inhibit ASK1. Although did not meet its primary end point change eGFR from baseline week 48, acute related inhibition creatinine secretion by confounded differences eGFR. Exploratory post hoc analyses accounting these suggest resulted reduction function decline merits further study. Background Apoptosis signal-regulating (ASK1) activation glomerular tubular cells resulting oxidative stress may drive progression. ASK1 inhibitor, Methods adults type diabetes treatment-refractory we randomly assigned 333 1:1:1:1 allocation (oral doses 2, 6, or 18 mg) placebo. Primary outcome was at weeks. Results Selonsertib mean placebo groups differ significantly Because this unanticipated effect, used piecewise linear regression, finding two effects: an more pronounced 0 4 weeks (creatinine effect) attenuated between (therapeutic higher selonsertib. A analysis (excluding data 20 sites Good Clinical Practice compliance–related issues) rate reduced 71% group relative (difference 3.11±1.53 ml/min per 1.73 m annualized year; 95% confidence interval, 0.10–6.13; nominal P =0.043). Effects on urine albumin-to-creatinine ratio Conclusions endpoint, exploratory slow

Language: Английский

Citations

87

Kidney Disease in Diabetic Patients: From Pathophysiology to Pharmacological Aspects with a Focus on Therapeutic Inertia DOI Open Access
Guido Gembillo, Ylenia Ingrasciotta, Salvatore Crisafulli

et al.

International Journal of Molecular Sciences, Journal Year: 2021, Volume and Issue: 22(9), P. 4824 - 4824

Published: May 1, 2021

Diabetes mellitus represents a growing concern, both for public economy and global health. In fact, it can lead to insidious macrovascular microvascular complications, impacting negatively on patients’ quality of life. Diabetic patients often present diabetic kidney disease (DKD), burdensome complication that be silent years. The average time onset impairment in is about 7–10 clinical impact DKD dangerous not only the risk progression end-stage renal therefore replacement therapies, but also because associated increase cardiovascular events. An early recognition factors decisive decreasing morbidity mortality. presents patient-related, clinician-related, system-related issues. All these problems are translated into therapeutic inertia, which defined as failure initiate or intensify therapy according evidence-based guidelines. Therapeutic inertia resolved by multidisciplinary pool healthcare experts. timing intensification treatment, transition best therapy, dietetic strategies must provided team, driving glycemic target delaying overcoming DKD-related complications. A timely nephrological evaluation guarantee adequate information choose right at case progression.

Language: Английский

Citations

62

Molecular Mechanisms of Diabetic Kidney Disease DOI Open Access
Jorge Rico-Fontalvo, Gustavo Aroca, José Cabrales

et al.

International Journal of Molecular Sciences, Journal Year: 2022, Volume and Issue: 23(15), P. 8668 - 8668

Published: Aug. 4, 2022

The inflammatory component of diabetic kidney disease has become great interest in recent years, with genetic and epigenetic variants playing a fundamental role the initiation progression disease. Cells innate immune system play major pathogenesis disease, lesser contribution from adaptive cells. Other components such as complement also role, well specific cytokines chemokines. is an active research field, hope to find potential innovative therapeutic targets.

Language: Английский

Citations

54

Precision prognostics for the development of complications in diabetes DOI Creative Commons
Catarina Schiborn, Matthias B. Schulze

Diabetologia, Journal Year: 2022, Volume and Issue: 65(11), P. 1867 - 1882

Published: June 21, 2022

Abstract Individuals with diabetes face higher risks for macro- and microvascular complications than their non-diabetic counterparts. The concept of precision medicine in aims to optimise treatment decisions individual patients reduce the risk major diabetic complications, including cardiovascular outcomes, retinopathy, nephropathy, neuropathy overall mortality. In this context, prognostic models can be used estimate an individual’s relevant based on profiles. This review place prediction modelling into context prognostics. As opposed identification subsets, development models, selection predictors longitudinal association outcome interest discriminatory ability, allows estimation absolute complications. a consequence, such provide information about potential patient subgroups needs. provides insight methodological issues specifically related validation We summarise existing commonly included predictors, examples available studies. also discusses non-classical markers omics-based predictors. Finally, it gives requirements challenges clinical applications implementation developed predictions medical decision making. Graphical abstract

Language: Английский

Citations

50