Blood Purification最新文献

Introduction: Patients with end-stage renal disease (ESRD) are known to have reduced structural and functional brain connectivity in the brain regions associated with cognitive function. However, the effect of dialysis on brain connectivity remains unclear. This study aimed to evaluate the effects of dialysis on structural brain connectivity in patients with ESRD.

Methods: This prospective study included 20 patients with ESRD in the pre-dialysis stage and 35 healthy controls. The patients underwent T2-weighted and three-dimensional T1-weighted magnetic resonance imaging before and 3 months after dialysis initiation. Moreover, the cortical thickness was calculated. We applied graph theoretical analysis to calculate the structural covariance network based on cortical thickness. We compared the cortical thickness and structural covariance network of patients with ESRD in the pre-dialysis stage with those of healthy controls and with those of patients with ESRD in the post-dialysis stage.

Results: The mean cortical thickness in both hemispheres was lower in patients with ESRD in the pre-dialysis stage than in healthy controls (2.296 vs. 2.354, p=0.030; 2.282 vs. 2.362, p=0.004, respectively) and was higher in patients with ESRD in the post-dialysis stage than in those in the pre-dialysis stage (2.333 vs. 2.296, p=0.001; 2.322 vs. 2.282, p=0.002, respectively). Analysis of the structural covariance network revealed that the assortative coefficient was lower in patients with ESRD in the pre-dialysis stage than in healthy controls (-0.062 vs. -0.031, p=0.029) and was higher in patients with ESRD in the post-dialysis stage than in those in the pre-dialysis stage (-0.002 vs. -0.062, p=0.042).

Conclusion: We observed differences in the cortical thickness and structural covariance networks before and after dialysis in patients with ESRD. This indicates that dialysis affects structural brain connectivity, contributing to the understanding of the pathophysiological mechanism of cognitive function alterations resulting from dialysis in patients with ESRD. .

Introduction: The effect of change in sodium (Na) in critically ill patients undergoing slow low-efficiency dialysis (SLED) is unclear.

Methods: Prospective observational study enrolled dysnatremic critically ill adult patients with acute kidney injury (AKI) undergoing the first SLED as cases and normonatremic patients as controls. Baseline and SLED-related parameters, and 30-day mortality were noted.

Results: 100 dysnatremic and 51 normonatremic patients were included, with a median age of 31(25-52) years and median admission SOFA scores of 10 (9-12). Patients with dysnatremia at study inclusion had a mortality of 53%, with the highest mortality in severe hypernatremia (Na > 160 mEq/L: 75%), followed by those with severe hyponatremia (Na < 120 mEq/L: 68.6%). SLED-associated natremia change > 10 mEq/L was significantly associated with mortality, in patients with mild dysnatremia and normonatremia (Na:130-150) (p<0.001), and not in those with moderate to severe dysnatremia (Na < 130 and Na >150) (p=0.72). Upon multivariate logistic regression analysis, a model with pre-SLED pH, dialysate-pre-SLED Na difference, and duration of SLED significantly predicted SLED-associated natremia change (R2 0.18, p=0.001).

Conclusions: Na change can be more than 10 meq/L in one-third of critically ill patients, subjected to first SLED session, which is associated with poor outcome in mild dysnatremics and normonatremics.

Introduction: Hypotension is common during intermittent hemodialysis (IHD) and may be due to a decreased cardiac index (CI). However, no study has simultaneously and continuously measured CI and mean arterial pressure (MAP) to understand the prevalence, severity, and duration of CI decreases or relate them to MAP, blood volume (BV) and net ultrafiltration (NUF) rate.

Methods: In a prospective, pilot and feasibility investigation, we studied 10 chronic IHD patients. We used the ClearSight System™ to continuously monitor CI and MAP; the CRIT-LINE®IV monitor to detect BV changes and collected data on NUF rate.

Results: Device tolerance and compliance was 100%. All patients experienced at least ≥ 1 episode of severe CI decrease (> 25% from baseline), with a median duration of 24 minutes [IQR 6-87] and of 68 minutes [14-106] for moderate decreases (>15% but  25% from baseline). Eight patients experienced a low CI state (<2.2 L/min/m2). The lowest CI was 0.9 L/min/m2 with a concomitant MAP of 94 mmHg. When the fall in CI was severe, MAP increased in 58% of cases and remained stable in 28%. Overall, CI decreased by -0.55 L/min/m2 when BV decrease was moderate vs mild (p<0.001) and by -0.8 L/min/m2 when NUF rate was high vs low (p<0.001).

Conclusion: Continuous CI monitoring is feasible in IHD and shows frequent moderate-severe CI decreases, sometimes to low CI state levels. Such decreases are typically associated with markers of decreased intravascular volume status but not with a decrease in MAP, implying marked vasoconstriction.

Introduction: Chronic kidney disease (CKD) poses a significant global health burden, with increasing prevalence and high morbidity and mortality rates, particularly in end-stage kidney disease (ESKD). While traditional risk factors contribute, the exact mechanisms remain elusive, with inflammation playing a pivotal role. Medium cut-off (MCO) membranes offer promise in improving dialysis outcomes by efficiently clearing uremic toxins without substantial albumin loss. We aimed to elucidate the impact of MCO and high-flux (HF) membranes on peripheral blood lymphocyte subpopulations in hemodialysis patients.

Methods: Twenty-four ESKD patients underwent 36 sessions each with MCO and HF membranes. Immunophenotyping by flow cytometry was performed to analyze lymphocyte subsets.

Results: NK cell percentages significantly increased with MCO, returning to baseline with HF. Th1 cells decreased post-HF, while Th2 and TFH cells increased with MCO and persisted. Treg cells remained stable with MCO but decreased with HF.

Conclusion: MCO dialysis induced an anti-inflammatory shift, evidenced by increased Th2 and TFH cells and stable Treg cells. NK cells also responded favorably to MCO. These findings underscore MCO membranes' potential to modulate immune responses and improve patient outcomes in ESKD.

Introduction: Chloride transfers during continuous renal replacement therapy (CRRT) have not been adequately described and may differ based on CRRT technique. We aimed to measure chloride mass transfer (JS,Cl) during CRRT and identify associated determinants.

Methods: We performed a two-center, prospective, observational study in France and Australia in ICU patients with CRRT initiated for < 24h. Patients received continuous veno-venous hemofiltration (CVVH) or continuous veno-venous hemodialysis (CVVHD, with citrate-CaCl2 regional anticoagulation). Over a 24h period, plasma and effluent chloride concentrations were measured every 4h to compute chloride mass transfer (JS,Cl, in mmol.min-1) using a modality-specific model, with negative value indicating chloride transfer towards the patient. Secondary outcomes were the identification of CRRT settings associated with JS,Cl (using multivariate mixed effects regression). Results are presented with median [interquartile range].

Results: Between February 2021 and August 2022, we enrolled 37 patients (64 [56-71] years, 67% male), for a total of 20 CVVHD and 20 CVVH sessions. Over 24h, plasma chloride concentrations were significantly higher, and JS,Cl significantly lower during CVVHD, compared to CVVH (-0.10 [-0.33-0.15] vs. 0.01 [-0.10-0.13] mmol.min-1, P<0.05). With both modalities, net ultrafiltration (QUFNET) and plasma chloride concentrations were the principal determinants of JS,Cl, with higher QUFNET being associated with an increase in JS,Cl during CVVHD. Also, CVVHD sessions demonstrated a concentration gradient between the plasma and the effluent chamber of -6 [-9- -4] mmol.L-1. Finally, CaCl2 reinjection during CVVHD accounted for 35% [32%-60%] of total JS,Cl in sessions with a negative JS,Cl.

Conclusion: Compared to CVVH, CVVHD with regional citrate anticoagulation was associated with greater chloride mass transfer to the patient and higher plasma chloride concentrations. This was due to high dialysate chloride concentrations and CaCl2 reinjection. This effect could only be controlled by high net ultrafiltration flow rates.

Background: Generative artificial intelligence (AI) is rapidly transforming various aspects of healthcare, including critical care nephrology. Large language models (LLMs), a key technology in generative AI, show promise in enhancing patient care, streamlining workflows, and advancing research in this field.

Summary: This review analyzes the current applications and future prospects of generative AI in critical care nephrology. Recent studies demonstrate the capabilities of LLMs in diagnostic accuracy, clinical reasoning, and continuous renal replacement therapy (CRRT) alarm troubleshooting. As we enter an era of multiagent models and automation, the integration of generative AI into critical care nephrology holds promise for improving patient care, optimizing clinical processes, and accelerating research. However, careful consideration of ethical implications and continued refinement of these technologies are essential for their responsible implementation in clinical practice. This review explores the current and potential applications of generative AI in nephrology, focusing on clinical decision support, patient education, research, and medical education. Additionally, we examine the challenges and limitations of AI implementation, such as privacy concerns, potential bias, and the necessity for human oversight.

Key messages: (i) LLMs have shown potential in enhancing diagnostic accuracy, clinical reasoning, and CRRT alarm troubleshooting in critical care nephrology. (ii) Generative AI offers promising applications in patient education, literature review, and academic writing within the field of nephrology. (iii) The integration of AI into electronic health records and clinical workflows presents both opportunities and challenges for improving patient care and research. (iv) Addressing ethical concerns, ensuring data privacy, and maintaining human oversight are crucial for the responsible implementation of AI in critical care nephrology.

Introduction: In this systematic review and meta-analysis, we aimed to review available data and provide pooled estimates of the predictive performance of urinary chemokine (CC motif) ligand (uCCL14') for persistent (≥48 hours) severe acute kidney injury (PS-AKI).

Methods: We searched MEDLINE, PubMed, Cochrane Library, and EMBASE for studies published up to April 11, 2023. We considered all studies including adults and reporting on the ability of uCCL14 to predict PS-AKI as defined by AKI persisting for 48 or 72 hours. Data extraction was performed by one investigator using a standardized form. It was checked for adequacy and completeness by another investigator.

Results: After screening, we identified 13 relevant studies. Among those, four (561 patients) provided sufficient data regarding the outcome of interest and were included. Considering each study cut-off value, pooled sensitivity and specificity were 0.85 (95% CI: 0.77- 0.90, I2 = 34.1%) and 0.96 (95% CI: 0.94 - 0.98, I2 = 53.7%) respectively. Pooled positive likelihood ratio (LR), negative LR, and diagnostic odds ratio were 8.98 (95% CI: 4.92 - 16.37, I2 = 23%), 0.25 (95% CI: 0.17 - 0.37, I2 = 0%) and 14.98 (95% CI: 3.55 - 63.27, I2 = 72.9%) respectively. The area under the curve estimated by summary receiver operating characteristics was 0.86 (95% CI: 0.70 - 0.95). Heterogeneity induced by the threshold effect was low (Spearman correlation coefficient: -0.30 p-value = 0.62) but significant for non-threshold effect. Risk of bias and concern for applicability according to the QUADAS-2 criteria was generally low. High risk in the index test due to the absence of prespecified thresholds was a concern for most studies.

Conclusion: Based on current evidence, uCCL14 appears to have a good predictive performance for the occurrence of PS-AKI. Interventional trials to study a biomarker-guided application of AKI care bundles and RRT are indicated.

Introduction: Hematocrit monitoring during continuous renal replacement therapy (CRRT) allows the continuous estimation of relative blood volume (RBV). This may enable early detection of intravascular volume depletion prior to clinical sequelae. We aimed to investigate the feasibility of extended RBV monitoring and its epidemiology during usual CRRT management by clinicians unaware of RBV. Moreover, we studied the association between changes in RBV and net ultrafiltration (NUF) rates.

Methods: In a cohort of adult intensive care unit patients receiving CRRT, we continuously monitored hematocrit and RBV using a pre-filter noninvasive optical sensor. We analyzed temporal changes in RBV and investigated the association between RBV change and NUF rates, using the classification of NUF rates into low, moderate, or high based on predefined cut-offs.

Results: We obtained >60,000 minute-by-minute measurements in >1,000 CRRT hours in 36 patients. The median RBV change was negative (decrease) in 69% of patients and the median peak change in RBV was -9.3% (interquartile range: -3.9% to -14.3%). Moreover, the median RBV decreased from baseline by >5% in 40.2% of measurements and by >10% in 20.6% of measurements. Finally, RBV decreased significantly more when patients received a high NUF rate (>1.75 mL/kg/h) compared to low or moderate NUF rates (5.32% vs. 1.93% or 1.97%, p < 0.001).

Conclusion: Continuous hematocrit and RBV monitoring during CRRT was feasible. RBV decreased significantly during CRRT, and decreases were greater with higher NUF rates. RBV monitoring may help optimize NUF management and prevent the occurrence of intravascular volume depletion.

Introduction: Continuous renal replacement therapy (CRRT) is a prolonged continuous extracorporeal blood purification therapy to replace impaired renal function. Typically, CRRT therapy requires routine anticoagulation, but for patients at risk of bleeding and with contraindications to sodium citrate, anticoagulant-free dialysis therapy is necessary. However, this approach increases the risk of CRRT circuit coagulation, leading to treatment interruption and increased resource consumption. In this study, we utilized artificial intelligence machine learning methods to predict the risk of CRRT circuit coagulation based on pre-CRRT treatment metrics.

Methods: We retrospectively analyzed 212 patients who underwent anticoagulant-free CRRT from October 2022 to October 2023. Patients were categorized into high-risk and low-risk groups based on CRRT circuit coagulation within 24 h. We employed eight machine learning methods to predict the risk of circuit coagulation. The performance of the model was evaluated using the area under the curve (AUC) of the receiver operating characteristic. 5-fold cross-validation was used to validate the machine learning models. Feature importance and SHAP plots were used to interpret the model's performance and key drivers.

Results: We identified 88 patients (41.51%) at high risk of circuit coagulation within 24 h of CRRT. Our machine learning models showed excellent predictive performance, with ensemble learning achieving an AUC of 0.863 (95% CI: 0.860-0.868), outperforming individual algorithms. Random forest was the best single-algorithm model, with an AUC of 0.819 (95% CI: 0.814-0.823). The top three features identified as most important by the SHAP summary plot and feature importance graph are platelet, filtration fraction (FF), and triglycerides.

Conclusion: We created a model using machine learning to predict the risk of circuit coagulation during anticoagulant-free CRRT therapy. Our model performs well (AUC 0.863) and identifies key factors like platelets, FF, and triglycerides. This facilitates the development of personalized treatment strategies by clinicians aimed at reducing circuit coagulation risk, thereby enhancing patient outcomes and reducing healthcare expenses.