Continuous renal replacement therapy (CRRT) is an important intervention in the treatment of acute kidney injury (AKI) in intensive care, and its application scope continues to expand alongside technological advancements and an increasing patient population. However, in current clinical practice, several factors, including individual patient differences, dynamic treatment needs, and the rational allocation of medical resources, pose challenges to CRRT implementation. The rapid development of artificial intelligence (AI) offers new approaches to address some of these challenges, and its applications in developing decision-support models, refining workflows, and supporting patient communication and education may contribute to more precise and efficient CRRT management. In this article, we discuss the current evidence and potential applications of AI in CRRT, as well as the challenges and strategies for clinical implementation, aiming to provide a reference for future research and clinical translation of AI-assisted CRRT.

Not applicable.

Acute‑on‑chronic kidney disease (AoCKD) refers to an acute decline in renal function occurring in individuals with established CKD. Despite its major clinical impact-including increased mortality, accelerated CKD progression, and higher cardiovascular risk-AoCKD remains underrepresented within major guideline frameworks (KDIGO, ADQI, NICE). Our review aims to critically appraise existing guideline frameworks, definitions, biomarkers, imaging recommendations, and research gaps related to AoCKD, and to highlight priorities for future harmonization and clinical integration. We conducted a targeted literature search of PubMed, Embase, and the Cochrane Library, to identify guidelines, consensus reports, observational studies, and systematic reviews describing acute deterioration in CKD. Extracted data included definitional criteria, epidemiology, biomarker performance, imaging guidance, and prognostic outcomes. Current KDIGO, ADQI, and NICE frameworks were evaluated in parallel. We report marked heterogeneity in definitions describing acute deterioration in CKD, with terms including AoCKD, AKI‑on‑CKD, acute pre‑existing kidney dysfunction, and cROCK‑based definitions. Reported incidence varies widely (1-33%) due to inconsistent criteria. Creatinine‑based definitions show reduced prognostic accuracy in CKD populations, whereas cystatin‑C-based eGFR demonstrates superior discrimination for mortality and progression. We propose an operational definition of AoCKD that includes functional criteria and biomarker-based criteria. The definition reflects the increased mortality in this phenotype even before reaching the KDIGO-AKI threshold. In conclusion, AoCKD is a clinically significant but poorly defined state within the kidney disease continuum. Harmonized definitions, incorporation of cystatin‑C-based markers, structured imaging pathways, and integration of early biomarkers are urgently needed.

Cross-elbow stent placement in hemodialysis arteriovenous fistulas (AVFs) has been avoided due to risks of fracture, kinking, deformation, migration, and infection. However, elbow lesions may be refractory to angioplasty and progress to chronic total occlusion (CTO). We evaluated the safety, feasibility, and patency of cross-elbow stenting for AVF salvage. Consecutive adult hemodialysis patients who received cross-elbow stents for clinically significant AVF stenosis or CTO (September 2021 - March 2026) at a single center were retrospectively analyzed. Primary outcomes were target lesion primary patency (TLPP), access circuit primary patency (ACPP), access circuit secondary patency, and access circuit assisted primary patency, estimated by Kaplan-Meier. Technical and clinical success and complications were recorded. Forty-two patients were included; 45.2% had CTO and 57.1% had prior PTA. Eluvia-based stents were used in 83.3%. Technical and clinical success were both 100%. Mean access flow increased from 407 ± 267 to 839 ± 188 mL/min. TLPP at 3, 6, 12, and 24 months was 90.1%, 84.4%, 76.9%, and 76.9%, respectively. No stent fracture, migration, kinking, or infection occurred. Cross-elbow stenting was technically feasible and provided favorable short- to mid-term patency without device-related complications, supporting its use as a salvage option for selected refractory AVF lesions and CTOs.

To compare the clinicopathological features, treatment response, and renal outcomes of patients with IgA nephropathy (IgAN) with κ-dominant, κ-λ co-deposition, or λ-dominant light-chain deposition. This retrospective, single-centre cohort study included 105 patients with primary IgAN who underwent their first native-kidney biopsy. Twenty-one patients with κ-dominant deposition were matched to 42 patients with κ-λ co-deposition and 42 patients with λ-dominant deposition based on age, sex, 24-hour urinary protein excretion, and baseline serum creatinine. The final matching ratio was 1:2:2. The primary renal endpoint was a sustained ≥30% increase in serum creatinine from baseline. During a median follow-up of 25 months (IQR, 18-36), 21 patients reached the renal endpoint. Time to treatment response did not differ significantly among the three groups (log-rank p = 0.0707). However, renal progression-free survival differed significantly among the light-chain deposition groups (log-rank p = 0.0046). After adjustment for sex, age, 24-hour urinary protein excretion, and baseline eGFR, a higher κ deposition ratio remained associated with a lower hazard of renal progression (HR [per 1-percentage-point increase] 0.96; 95% CI: 0.94-0.99; p = 0.011). A higher κ light-chain deposition ratio was associated with a lower risk of renal progression in IgAN. κ predominance may be a pathological marker of a distinct and possibly less aggressive immune-complex deposition pattern.

Maternal chronic kidney disease (CKD) is associated with adverse pregnancy outcomes, but its relationship with long-term offspring kidney health remains uncertain. This retrospective maternal-child cohort study examined whether maternal CKD or reduced kidney function before or during pregnancy was associated with a composite kidney outcome in offspring. Live births in the Taiwan Mother and Child Health cohort from 2007 to 2017 were followed through 2018. In the primary cohort, maternal kidney disease was defined as claims-based CKD meeting chronicity criteria before or during pregnancy; the maternal eGFR subcohort additionally included eGFR-defined reduced kidney function. The primary outcome comprised ICD-defined CKD, reduced eGFR <60 mL/min/1.73 m[2] or ≥30% eGFR decline at age ≥2 years, or kidney replacement therapy. Laboratory components were interpreted as reduced kidney function rather than definitive CKD unless diagnostic chronicity was documented. Models adjusted for maternal, pregnancy, offspring, and perinatal covariates. Among 79,036 children, 4,687 (5.9%) met the composite kidney outcome. Maternal kidney disease was associated with modestly higher odds of the composite outcome (aOR, 1.29; 95% CI, 1.00-1.66) and offspring all-cause mortality (aOR, 1.88; 95% CI, 1.15-3.09). Analyses of individual components of the composite outcome, sex-stratified subgroups, and trimester-specific exposure windows were exploratory, imprecise, and not adjusted for multiplicity. Maternal CKD or reduced kidney function may identify dyads who warrant long-term kidney surveillance, although causal pathways require confirmation.

This study investigated the association between triglyceride-to-high-density lipoprotein cholesterol ratio (TG/HDL-C) and kidney stone status and evaluated its contribution to multivariable classification models. A total of 1,174 patients from the First Affiliated Hospital of Anhui Medical University were included, with kidney stone status confirmed by urological imaging. Multivariable analyses were used to assess the association between TG/HDL-C and kidney stone status. Supplementary population-based analyses were conducted using weighted data from the National Health and Nutrition Examination Survey (NHANES). Seven machine-learning models were developed and internally evaluated within the clinical cohort, and SHapley Additive Explanations (SHAP) were used to interpret model outputs. In the clinical cohort, a higher TG/HDL-C ratio was associated with greater odds of kidney stone status after multivariable adjustment (OR, 2.16; 95% CI, 1.73-2.70). The association was generally consistent across subgroups defined by age, sex, body mass index, hypertension, and diabetes, with no significant interactions observed. A directionally similar association was identified in the weighted NHANES analysis (OR, 1.04; 95% CI, 1.01-1.07). Among the evaluated machine-learning models, the random forest model achieved the highest area under the receiver operating characteristic curve in the internal validation set (AUROC, 0.818; 95% CI, 0.775-0.859). SHAP analysis identified the TG/HDL-C ratio as an important contributor to model classification. These findings indicate that the TG/HDL-C ratio is an accessible metabolic marker associated with kidney stone status. However, the observational design precludes causal inference, and prospective studies with external model validation are warranted.