Adrià Juanola

Background & Aims: Alcohol-related hepatitis (AH) is a severe form of alcohol-related liver disease (ALD). The cellular subpopulations that contribute to AH and dynamics remain poorly understood. This study aimed to identify key cellular subpopulations and their interactions in AH and explore key mechanisms contributing to disease pathogenesis. Approach & Results: Single-cell RNA sequencing was performed starting from liver biopsies from patients with AH, early-stage ALD, and healthy individuals. Immunofluorescence was used to validate protein expression of selected markers. Liver biopsy-derived organoids were generated and exposed to a conditioned medium recapitulating AH (AH-medium) to assess serum amyloid A (SAA) secretion by ELISA. Circulating SAA levels were measured in plasma samples obtained from patients at different stages of ALD. Our analysis revealed an atlas of several cell subpopulations of AH, including hepatocytes expressing SAA, scar-associated (SA) endothelial cells, Fibulin-3 ⁺ mesenchymal cells, SA-macrophages expressing MMP19, activated CD4+ and stressed/effector CD8+ T cell subsets, and hyperactivated neutrophils expressing FPR2. Cell-communication analysis predicted a strong SAA-FPR2 interaction between hepatocyte and neutrophil subpopulations in AH. We confirmed that AH-hepatocytes secreted SAA in organoid cultures after stimulation with AH-medium. We also found out that circulating SAA levels were elevated in AH and were associated with reduced survival. Conclusions: The SAA-FPR2 pathway emerges as a potential mediator of inflammation in AH, offering potential therapeutic targets for this severe disease. This study provides new insights into the cellular heterogeneity and dynamics in ALD progression, dissecting AH at a single-cell level.

In response to the growing health crisis of liver-related morbidity and mortality, screening for liver cirrhosis has emerged as a promising strategy for early detection and timely intervention. By identifying individuals with severe fibrosis or compensated cirrhosis, screening holds the promise of enhancing treatment outcomes, delaying disease progression, and ultimately improving the quality of life of affected individuals. Clinical practice guidelines from international scientific societies currently recommend targeted screening strategies, investigating high-risk populations with known risk factors of liver disease. While there is good evidence that screening increases case finding in the population, and a growing number of studies indicate that screening may motivate beneficial lifestyle changes in patients with steatotic liver disease, there are major gaps in knowledge in need of clarification before screening programs of cirrhosis are implemented. Foremost, randomized trials are needed to ensure that screening leads to improved liver-related morbidity and mortality. If not, screening for cirrhosis could be unethical due to overdiagnosis, overtreatment, increased health care costs, negative psychological consequences of screening, and futile invasive investigations. Moreover, the tests used for screening need to be optimized toward lower false positive rates than the currently used FIB-4 while retaining few false negatives. Finally, barriers to adherence to screening and implementation of screening programs need to be elucidated. This review provides a comprehensive overview of the current landscape of screening strategies for liver cirrhosis and the promises and pitfalls of current methods for early cirrhosis detection.

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.

An abstract is unavailable. This article is available as a PDF only.

Background Patients with decompensated cirrhosis experience high mortality rates. Current prognostic scores, including the model for end-stage liver disease (MELD), may underperform in settings other than in those they were initially developed. Novel biomarkers have been proposed to improve prognostication accuracy and even to predict development of complications. Methods We performed a systematic review and meta-analysis on novel urine and blood biomarkers and their ability to predict 90-day mortality in patients with decompensated cirrhosis. Secondary outcomes included 28-day and 1-year mortality, and development of acute-on-chronic liver failure, acute kidney injury and other complications. To overcome differences in units, temporal changes in assays and reporting heterogeneity, we used the ratio of means (RoM) as measure of association for assessing strength in predicting outcomes. An RoM>1 implies that the mean biomarker level is higher in those that develop the outcome than in those that do not. Results Of 6629 unique references, 103 were included, reporting on 29 different biomarkers, with a total of 31 362 biomarker patients. Most studies were prospective cohorts of hospitalised patients (median Child-Pugh-Turcotte score of 9 and MELD score of 18). The pooled 90-day mortality rate was 0.27 (95% CI 0.24 to 0.29). The RoM for predicting 90-day mortality was highest for interleukin 6 (IL-6) (2.56, 95% CI 2.39 to 2.74), followed by urinary neutrophil gelatinase-associated lipocalin (uNGAL) (2.42, 95% CI 2.20 to 2.66) and copeptin (2.33, 95% CI 2.17 to 2.50). These RoMs were all higher than for MELD (1.44, 95% CI 1.42 to 1.46). Conclusion Novel biomarkers, including IL-6, uNGAL and copeptin, can probably improve prognostication of patients with decompensated cirrhosis compared with MELD alone.

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.

Kidney biomarkers appear to be useful in differential diagnosis between acute tubular necrosis (ATN) and other types of acute kidney injury (AKI) in cirrhosis, particularly hepatorenal syndrome (HRS‐AKI). Distinction is important because treatment is different. However, kidney biomarkers are still not used in clinical practice. The aim of the current study was to investigate the accuracy of several biomarkers in differential diagnosis of AKI and in predicting kidney outcome and patient survival. This was a prospective study of 320 consecutive cases of AKI in patients hospitalized for decompensated cirrhosis. Evaluation of AKI was made with a diagnostic algorithm that included identification and removal/treatment of precipitating factors and albumin administration (1 g/kg for 2 days) to patients with AKI stage 1B or greater. Urinary neutrophil gelatinase–associated lipocalin (NGAL), monomeric NGAL (mNGAL), interleukin‐18, and standard biomarkers were measured at diagnosis and on days 3, 7, and 14. Of the 320 cases, 153 were hypovolemia‐induced AKI (48%), 93 were HRS‐AKI (29%), 39 were ATN (12%), and 35 were due to miscellaneous causes (11%). Among all biomarkers, urinary NGAL measured at day 3 had the greatest accuracy for differential diagnosis between ATN and other types of AKI (area under the receiver operating characteristic curve, 0.87; 95% confidence interval, 0.78‐0.95). The cutoff with the best predictive accuracy for ATN diagnosis was 220 µg/g creatinine. Progression of AKI during hospitalization was associated with persistently high NGAL levels, and NGAL was an independent predictive factor of AKI progression. Likewise, NGAL was also an independent predictive factor of 28‐day mortality together with Model for End‐Stage Liver Disease score. Conclusion: These results support the use of NGAL in clinical practice within the context of a diagnostic algorithm for differential diagnosis of AKI and outcome prediction in cirrhosis.