ABSTRACT
Acute kidney injury (AKI) is a common complication among patients with decompensated cirrhosis and its development is associated with worse prognosis in terms of survival. Patients with decompensated cirrhosis may develop a unique type of AKI, known as hepatorenal syndrome (HRS-AKI), characterized by marked impairment of kidney function due to haemodynamic changes that occur in late stages of liver cirrhosis. Besides, patients with cirrhosis also may develop chronic alterations of kidney function (chronic kidney disease, CKD), the incidence of which is increasing markedly and may be associated with clinical complications. The aim of this review is to provide the reader with an update of the most relevant aspects of alterations of kidney function in patients with cirrhossi that may be useful for theri clinical practice.
Subject(s)
Acute Kidney Injury , Hepatorenal Syndrome , Hypertension, Portal , Liver Cirrhosis , Humans , Hypertension, Portal/etiology , Hypertension, Portal/physiopathology , Hypertension, Portal/complications , Hepatorenal Syndrome/etiology , Hepatorenal Syndrome/physiopathology , Acute Kidney Injury/etiology , Acute Kidney Injury/physiopathology , Liver Cirrhosis/complications , Liver Cirrhosis/physiopathology , Renal Insufficiency, Chronic/complications , Renal Insufficiency, Chronic/physiopathologyABSTRACT
BACKGROUND: Central collection of distributed medical patient data is problematic due to strict privacy regulations. Especially in clinical environments, such as clinical time-to-event studies, large sample sizes are critical but usually not available at a single institution. It has been shown recently that federated learning, combined with privacy-enhancing technologies, is an excellent and privacy-preserving alternative to data sharing. OBJECTIVE: This study aims to develop and validate a privacy-preserving, federated survival support vector machine (SVM) and make it accessible for researchers to perform cross-institutional time-to-event analyses. METHODS: We extended the survival SVM algorithm to be applicable in federated environments. We further implemented it as a FeatureCloud app, enabling it to run in the federated infrastructure provided by the FeatureCloud platform. Finally, we evaluated our algorithm on 3 benchmark data sets, a large sample size synthetic data set, and a real-world microbiome data set and compared the results to the corresponding central method. RESULTS: Our federated survival SVM produces highly similar results to the centralized model on all data sets. The maximal difference between the model weights of the central model and the federated model was only 0.001, and the mean difference over all data sets was 0.0002. We further show that by including more data in the analysis through federated learning, predictions are more accurate even in the presence of site-dependent batch effects. CONCLUSIONS: The federated survival SVM extends the palette of federated time-to-event analysis methods by a robust machine learning approach. To our knowledge, the implemented FeatureCloud app is the first publicly available implementation of a federated survival SVM, is freely accessible for all kinds of researchers, and can be directly used within the FeatureCloud platform.
ABSTRACT
BACKGROUND & AIMS: The management of acute kidney injury (AKI) in cirrhosis is challenging. The EASL guidelines proposed an algorithm for the management of AKI, but this has never been validated. We aimed to prospectively evaluate this algorithm in clinical practice. METHODS: We performed a prospective cohort study in consecutive hospitalized patients with cirrhosis and AKI. The EASL management algorithm includes identification/treatment of precipitating factors, 2-day albumin infusion in patients with AKI ≥stage 1B, and treatment with terlipressin in patients with hepatorenal syndrome (HRS-AKI). The primary outcome was treatment response, which included both full and partial response. Secondary outcomes were survival and adverse events associated with terlipressin therapy. RESULTS: A total of 202 AKI episodes in 139 patients were included. Overall treatment response was 80%, while renal replacement therapy was required in only 8%. Response to albumin infusion was achieved in one-third of episodes. Of patients not responding to albumin, most (74%) did not meet the diagnostic criteria of HRS-AKI, with acute tubular necrosis (ATN) being the most common phenotype. The response rate in patients not meeting the criteria for HRS-AKI was 70%. Only 30 patients met the diagnostic criteria for HRS-AKI, and their response rate to terlipressin was 61%. Median time from AKI diagnosis to terlipressin initiation was only 2.5 days. While uNGAL (urinary neutrophil gelatinase-associated lipocalin) could differentiate ATN from other phenotypes (AUROC 0.78), it did not predict response to therapy in HRS-AKI. Ninety-day transplant-free survival was negatively associated with MELD-Na, ATN and HRS-AKI as well as uNGAL. Three patients treated with terlipressin developed pulmonary edema. CONCLUSIONS: The application of the EASL AKI algorithm is associated with very good response rates and does not significantly delay initiation of terlipressin therapy. IMPACT AND IMPLICATIONS: The occurrence of acute kidney injury (AKI) in patients with cirrhosis is associated with poor short-term mortality. Improving its rapid identification and prompt management was the focus of the recently proposed EASL AKI algorithm. This is the first prospective study demonstrating that high AKI response rates are achieved with the use of this algorithm, which includes identification of AKI, treatment of precipitating factors, a 2-day albumin challenge in patients with AKI ≥1B, and supportive therapy in patients with persistent AKI not meeting HRS-AKI criteria or terlipressin with albumin in those with HRS-AKI. These findings support the use of this algorithm in clinical practice.
Subject(s)
Antibodies, Monoclonal, Humanized , Antineoplastic Agents, Immunological , Chemical and Drug Induced Liver Injury , Humans , Antibodies, Monoclonal, Humanized/adverse effects , Chemical and Drug Induced Liver Injury/etiology , Chemical and Drug Induced Liver Injury/immunology , Chemical and Drug Induced Liver Injury/diagnosis , Antineoplastic Agents, Immunological/adverse effects , Male , Melanoma/drug therapy , Melanoma/immunology , Female , Immune Checkpoint Inhibitors/adverse effects , Liver/pathology , Liver/drug effects , Liver/immunology , Middle Aged , AgedABSTRACT
INTRODUCTION: Human albumin is used in the treatment of complications of cirrhosis. However, the use of long-term human albumin administration is costly and resource demanding for both patients and healthcare systems. A precision medicine approach with biomarkers to predict human albumin treatment response, so-called predictive biomarkers, could make this a viable treatment option in patients with cirrhosis and ascites. METHODS AND ANALYSIS: ALB-TRIAL is a multinational, double-blind, placebo-controlled randomised controlled trial. We aim to validate a predictive biomarker, consisting of a panel of circulating metabolites, to predict the treatment response to human albumin in patients with cirrhosis and ascites. All enrolled patients are stratified into a high-expected or low-expected effect stratum of human albumin based on the biomarker outcome. After stratification, patients in each group are randomised into either active treatment (20% human albumin) or corresponding placebo (0.9% NaCl) every 10th day for 6 months. The primary outcome is the cumulative number of liver-related events (composite of decompensation episodes, transjugular intrahepatic shunt insertion, liver transplantation and death). Key secondary outcomes include time-to-event analysis of primary outcome components, an analysis of the total healthcare burden and a health economic analysis. ETHICS AND DISSEMINATION: The trial obtained ethical and regulatory approval in Denmark, Germany, the Netherlands, Belgium, Hungary and Spain through the Clinical Trials Information System (CTIS) from 13 February 2023, while UK approvals from the Health Regulatory Authority, Medicines and Healthcare products Regulatory Agency and Research Ethics Committee are pending. Findings will be published in peer-reviewed journals, presented at conferences, communicated to relevant stakeholders and in the public registry of CTIS, following trial completion. TRIAL REGISTRATION NUMBER: NCT05056220 EU CT: 2022-501006-34-01.
Subject(s)
Liver Transplantation , Serum Albumin, Human , Humans , Serum Albumin, Human/therapeutic use , Ascites/therapy , Liver Cirrhosis/complications , Treatment Outcome , Biomarkers , Double-Blind MethodABSTRACT
Hepatorenal syndrome (HRS) is a form of kidney dysfunction that characteristically occurs in liver cirrhosis. It is characterized by a marked impairment of kidney function in response to circulatory and hemodynamic alterations that occur in advanced stages of liver cirrhosis, aggravated by systemic inflammation and bacterial translocation. The classical definitions of the types of HRS have been recently revisited and 2 forms of HRS have been redefined: the acute form, referred to as acute kidney injury (HRS-AKI), and the chronic form, referred to as chronic kidney disease. HRS-AKI is one of the most severe forms of AKI in patients with cirrhosis and it consists of an abrupt impairment of kidney function, frequently triggered by an infection, appearing in the setting of advanced decompensated cirrhosis. Differential diagnosis with other causes of AKI is crucial because HRS-AKI requires a specific treatment. Differential diagnosis with AKI-acute tubular necrosis may be challenging and kidney biomarkers may be useful in this setting. Treatment of HRS-AKI is based on the administration of vasoconstrictor drugs in combination with volume expansion with albumin. Prognosis of HRS-AKI is poor, and the ideal definitive treatment consists of liver transplantation or simultaneous liver-kidney transplantation. HRS-AKI has a big impact on patients' quality of life. Management of HRS-AKI remains challenging in specific situations such as alcohol-associated hepatitis or metabolic-associated steatotic liver disease cirrhosis. Developing preventive measures for HRS-AKI, improving its early identification, discovering new biomarkers for differential diagnosis, and improving the response to therapy are some of the unmet needs in the field of HRS-AKI.