Use of Liver Imaging and Biopsy in Clinical Practice

Incidence and Predictors of Complications Following Percutaneous Liver Biopsy: A Large Italian Multicentre Study

BACKGROUND AND AIMS

Management of ultrasound (US)-guided percutaneous liver biopsy (PLB) lacks standardisation. Despite the low risk of major complications (< 1%), repeated blood counts and up to 4 h of in-hospital observation are typically recommended. We aimed to assess complication rates and predictors in a large cohort of patients undergoing US-PLB across three Italian tertiary centres. We included all patients undergoing US-PLB from January 2018 to December 2023. We collected clinical, biochemical and procedural features (needle type, insertions number, observation time and antiplatelet/anticoagulant regimens). Safety was assessed by the incidence of pain and major complications, including vasovagal reaction, bleeding, pneumothorax, shock, hospitalisation and death.

APPROACH AND RESULTS

Among 1838 patients (mean age 55.1 years, 46.1% females, mean BMI 25.1 kg/m2, 74% parenchymal PLB), few were on anticoagulant/antiplatelet therapy (4.2%/16.2%); mean platelet count and PT INR were 209.7 × 103/mm3 and 1.04; 17 patients (0.9%) received prophylactic therapy (blood components, thrombopoietin receptor agonists or vitamin K). Needle aspiration was predominant (92%). During a mean time of observation of 5.4 ± 2.0 h, 134/1838 patients reported pain (7.4%). Major complications were few (26, 1.4%), with 14 episodes of transient hypotension (0.8%); 10 bleeding events (0.5%); 1 pneumothorax; 1 abscess formation; 1 haemobilia; and 1 episode of chest pain. Hospitalisation was rare (12, 0.7%), primarily for bleeding management. No fatalities were recorded. Pain was the sole significant independent predictor of major complications (18/26, HR 30.6, p < 0.05), particularly when reported within the first hour post-procedure (15/18, 83.3%).

CONCLUSIONS

Major complications following US-PLB are few and strongly associated with early post-procedural pain. In the absence of pain within the first hour, extended monitoring may be unnecessary, allowing for earlier patient discharge.

Enhancing Liver Nodule Visibility and Diagnostic Classification Using Ultrasound Local Attenuation Coefficient Slope Imaging

OBJECTIVE

B-mode ultrasound (US) presents challenges in accurately detecting and distinguishing between benign and malignant liver nodules. This study utilized quantitative US local attenuation coefficient slope (LACS) imaging to address these limitations.

MATERIALS AND METHODS

This is a prospective, cross-sectional study in adult patients with definable solid liver nodules at US conducted from March 2021 to December 2023. The composite reference standard included histopathology when available or magnetic resonance imaging. LACS images were obtained using a phantom-free method. Nodule visibility was assessed by computing the contrast-to-noise ratio (CNR). Classification accuracy for differentiating benign and malignant lesions was assessed with the area under the receiver operating characteristic curve (AUC), along with sensitivity and specificity.

RESULTS

The study enrolled 97 patients (age: 62 y ± 13 [standard deviation]), with 57.0% malignant and 43.0% benign observations (size: 26.3 ± 18.9 mm). LACS images demonstrated higher CNR (12.3 dB) compared to B-mode (p < 0.0001). The AUC for differentiating nodules and liver parenchyma was 0.85 (95% confidence interval [CI]: 0.79-0.90), with higher values for malignant (0.93, CI: 0.88-0.97) than benign nodules (0.76, CI: 0.66-0.87). A LACS threshold of 0.94 dB/cm/MHz provided a sensitivity of 0.83 (CI: 0.74-0.89) and a specificity of 0.82 (CI: 0.73-0.88). LACS mean values were higher (p < 0.0001) in malignant (1.28 ± 0.27 dB/cm/MHz) than benign nodules (0.98 ± 0.19 dB/cm/MHz).

CONCLUSION

LACS imaging improves nodule visibility and provides better differentiation between benign and malignant liver nodules, showing promise as a diagnostic tool.

Human genetics of metabolic dysfunction-associated steatotic liver disease: from variants to cause to precision treatment

Metabolic dysfunction-associated steatotic liver disease (MASLD) is characterized by increased hepatic steatosis with cardiometabolic disease and is a leading cause of advanced liver disease. We review here the genetic basis of MASLD. The genetic variants most consistently associated with hepatic steatosis implicate genes involved in lipoprotein input or output, glucose metabolism, adiposity/fat distribution, insulin resistance, or mitochondrial/ER biology. The distinct mechanisms by which these variants promote hepatic steatosis result in distinct effects on cardiometabolic disease that may be best suited to precision medicine. Recent work on gene-environment interactions has shown that genetic risk is not fixed and may be exacerbated or attenuated by modifiable (diet, exercise, alcohol intake) and nonmodifiable environmental risk factors. Some steatosis-associated variants, notably those in patatin-like phospholipase domain-containing 3 (PNPLA3) and transmembrane 6 superfamily member 2 (TM6SF2), are associated with risk of developing adverse liver-related outcomes and provide information beyond clinical risk stratification tools, especially in individuals at intermediate to high risk for disease. Future work to better characterize disease heterogeneity by combining genetics with clinical risk factors to holistically predict risk and develop therapies based on genetic risk is required.

Transfer learning radiomic model predicts intratumoral tertiary lymphoid structures in hepatocellular carcinoma: a multicenter study

BACKGROUND

Intratumoral tertiary lymphoid structures (iTLS) in hepatocellular carcinoma (HCC) are associated with improved survival and may influence treatment decisions. However, their non-invasive detection remains challenging in HCC. We aim to develop a non-invasive model using baseline contrast-enhanced MRI to predict the iTLS status.

METHODS

A total of 660 patients with HCC who underwent surgery were retrospectively recruited from four centers between October 2015 and January 2023 and divided into training, internal test, and external validation sets. After features dimensionality and selection, corresponding features were used to construct transfer learning radiomic (TLR) models for diagnosing iTLS, and model interpretability was explored with pathway analysis in The Cancer Genome Atlas-Liver HCC. The performances of models were assessed using the area under the receiver operating characteristic curve (AUC). The log-rank test was used to evaluate the prognostic value of the TLR model. The combination therapy set of 101 patients with advanced HCC treated with first-line anti-programmed death 1 or ligand 1 plus antiangiogenic treatment between January 2021 and January 2024 was used to investigate the value of the TLR model for evaluating the treatment response.

RESULTS

The presence of iTLS was identified in 46.0% (n=308) patients. The TLR model demonstrated excellent performance in predicting the presence of iTLS in training (AUC=0.91, 95% CI: 0.87, 0.94), internal test (AUC=0.85, 95% CI: 0.77, 0.93) and external validation set (AUC=0.85, 95% CI: 0.81, 0.90). The TLR model-predicted iTLS group has favorable overall survival (HR=0.66; 95% CI: 0.48, 0.90; p=0.007) and relapse-free survival (HR=0.64; 95% CI: 0.48, 0.85; p=0.001) in the external validation set. The model-predicted iTLS status was associated with inflammatory response and specific tumor-associated signaling activation (all p<0.001). The proportion of treatment responders was significantly higher in the model-predicted group with iTLS than in the group without iTLS (36% vs 13.73%, p=0.009).

CONCLUSION

The TLR model has indicated accurate prediction of iTLS status, which may assist in the risk stratification for patients with HCC in clinical practice.

Photon-counting CT-derived Quantification of Hepatic Fat Fraction: A Clinical Validation Study

Background Steatosis is a critical health problem, creating a growing need for opportunistic screening. Early detection may allow for effective treatment and prevention of further liver complications. Purpose To evaluate photon-counting CT (PCCT) fat quantification on contrast-enhanced scans and validate the results against fat quantification via histopathologic assessment, controlled attenuation parameter (CAP) from transient elastography, and MRI proton density fat fraction (PDFF). Materials and Methods In this prospective, observational clinical study, PCCT-derived fat fraction quantification was assessed in participants with known or suspected liver disease. Participants underwent PCCT between February 2022 and January 2024. Participants also underwent biopsy, US with CAP measurement, or MRI with a PDFF sequence for hepatic fat fraction quantification. Liver fat fraction was measured on virtual noncontrast PCCT images using spectral processing software with a three-material decomposition algorithm for fat, liver tissue, and iodine. Steatosis was graded for each modality. Correlation between PCCT-based steatosis grades and biopsy- and CAP-based grades was assessed with the Spearman correlation coefficient. Agreement between PCCT and MRI PDFF measurements was assessed with the intraclass correlation coefficient. Receiver operating characteristic curve analysis was conducted to determine the optimal PCCT fat fraction threshold for distinguishing between participants with and those without steatosis. Results The study included 178 participants, of whom 27 (mean age, 60.7 years ± 15.2 [SD]; 18 male participants) underwent liver biopsy, 26 (mean age, 60.0 years ± 18.3; 15 male participants) underwent CAP measurement, and 125 (mean age, 61.2 years ± 13.1; 70 male participants) underwent MRI PDFF measurement. There was excellent agreement between PCCT and MRI PDFF assessment of liver fat fraction (intraclass correlation coefficient, 0.91 [95% CI: 0.87, 0.94]). In stratified analysis, the intraclass correlation coefficient was 0.84 (95% CI: 0.63, 0.93) in participants with known fibrosis and 0.92 (95% CI: 0.88, 0.94) in participants without fibrosis. There was moderate correlation of PCCT-based steatosis grade with histologic (ρ = 0.65) and CAP-based (ρ = 0.45) steatosis grade. Based on the Youden index, the PCCT fat fraction threshold that best discriminated between participants with and those without steatosis was 4.8%, with a maximum achievable sensitivity of 81% (38 of 47) and a specificity of 71% (55 of 78). Conclusion PCCT in a standard clinical setting allowed for accurate estimation of liver fat fraction compared with MRI PDFF-based reference standard measurements. © RSNA, 2025 See also the editorial by Kartalis and Grigoriadis in this issue.

The Liver Cirrhosis Network Cohort Study: Cirrhosis Definition, Study Population, and Endpoints

INTRODUCTION

One of the primary goals of the Liver Cirrhosis Network (LCN) is to develop a cohort study to better understand and predict the risk of hepatic decompensation and other clinical and patient-reported outcomes among patients with Child A cirrhosis.

METHODS

The LCN consists of a Scientific Data Coordinating Center and 10 clinical centers whose investigators populate multiple committees. The LCN Definitions and Measurements Committee developed preliminary definitions of cirrhosis and its complications by literature review, expert opinion, and reviewing definition documents developed by other organizations. The Cohort Committee developed the study protocol with the input of the steering committee.

RESULTS

The LCN developed a prospective cohort study to describe and predict the rates of incident clinical events pertaining to first decompensation and patient-reported outcomes. The LCN developed a pragmatic definition of compensated cirrhosis incorporating clinical, laboratory, imaging, and histological criteria. Definitions of incident and recompensated ascites, overt hepatic encephalopathy, variceal hemorrhage, bleeding because of portal gastropathy, and hepatocellular carcinoma were also codified.

DISCUSSION

The LCN Cohort Study design will inform the natural history of cirrhosis in contemporary patients with compensated cirrhosis. The LCN Definitions and Measures Committee developed criteria for the definition of cirrhosis to standardize entry into this multicenter cohort study and standardized criteria for liver-related outcome measures. This effort has produced definitions intended to be both sensitive and specific as well as easily operationalized by study staff such that outcomes critical to the LCN cohort are identified and reported in an accurate and generalizable fashion.

REGISTRATION

NCT05740358.

Integrating PNPLA3 into clinical risk prediction

The PNPLA3-rs738409-G variant was the first common variant associated with hepatic fat accumulation and progression of metabolic dysfunction-associated steatotic liver disease (MASLD). Nevertheless, to date, the clinical translation of this discovery has been minimal because it has not yet been clearly demonstrated where the genetic information may play an independent and additional role in clinical risk prediction. In this mini-review, we will discuss the most relevant evidence regarding the potential integration of the PNPLA3 variant into scores and algorithms for liver disease diagnostics and risk stratification, specifically focusing on MASLD but also extending to liver diseases of other etiologies. The PNPLA3 variant adds little in diagnosing the current state of the disease, whether in terms of presence/absence of metabolic dysfunction-associated steatohepatitis or the stage of fibrosis. While it can play an important role in prediction, allowing for the early definition of risk profiles that enable tailored monitoring and interventions over time, this is most valuable when applied to populations with relatively high pre-test probability of having significant fibrosis based on either non-invasive tests (e.g. Fibrosis-4) or demographics (e.g. diabetes). Indeed, in this context, integrating FIB4 with the PNPLA3 genotype can refine risk stratification, though there is still no evidence that genetic information adds to liver stiffness determined by elastography. Similarly, in patients with known liver cirrhosis, knowing the PNPLA3 genotype can play a role in predicting the risk of hepatocellular carcinoma, while more doubts remain about the risk of decompensation.

Comparative Analysis of the Total Proteome in Nonalcoholic Steatohepatitis: Identification of Potential Biomarkers

Nonalcoholic fatty liver disease is a hepatic condition characterized by excessive fat accumulation in the liver with advanced stage nonalcoholic steatohepatitis (NASH), potentially leading to liver fibrosis, cirrhosis, and cancer. Currently, the identification and classification of NASH require invasive liver biopsy, which has certain limitations. Mass spectrometry-based proteomics can detect crucial proteins and pathways implicated in NASH development and progression. We collected the liver and serum samples from choline-deficient, L-amino acid-defined high-fat diet fed NASH C57BL/6J mice and human serum samples to examine proteomic alterations and identify early biomarkers for NASH diagnosis. In-depth targeted multiple reaction monitoring scanning and immunoblotting assays were used to verify the biomarker candidates from mouse liver and serum samples, and enzyme-linked immunosorbent assay (ELISA) was employed to analyze human serum samples. The multiple reaction monitoring analysis of NASH liver revealed 50 proteins with altered expression (21 upregulated and 29 downregulated) that are involved in biological processes such as detoxification, fibrosis, inflammation, and fatty acid metabolism. Ingenuity pathway analysis identified impaired protein synthesis, cellular stress and defense, cellular processes and communication, and metabolism in NASH mouse liver. Immunoblotting analysis confirmed that the expression of proteins associated with fatty acid metabolism (Aldo B and Fasn) and urea cycle (Arg1, Cps1, and Otc) was altered in the mouse liver and serum. Further analysis on human serum samples using ELISA confirmed the increased expression of multiple proteins, including Aldo B, Asl, and Lgals3, demonstrating values of 0.917, 0.979, and 0.965 of area under the curve in NASH diagnosis. These findings offer valuable insights into the molecular mechanisms of NASH and possible diagnostic biomarkers for early detection.

Application of Interpretable Machine Learning Models to Predict the Risk Factors of HBV-Related Liver Cirrhosis in CHB Patients Based on Routine Clinical Data: A Retrospective Cohort Study

Chronic hepatitis B (CHB) infection represents a significant global public health issue, often leading to hepatitis B virus (HBV)-related liver cirrhosis (HBV-LC) with poor prognoses. Early identification of HBV-LC risk is essential for timely intervention. This study develops and compares nine machine learning (ML) models to predict HBV-LC risk in CHB patients using routine clinical and laboratory data. A retrospective analysis was conducted involving 777 CHB patients, with 50.45% (392/777) progressing to HBV-LC. Admission data consisted of 52 clinical and laboratory variables, with missing values addressed using multiple imputation. Feature selection utilized Least Absolute Shrinkage and Selection Operator (LASSO) regression and the Boruta algorithm, identifying 24 key variables. The evaluated ML models included XGBoost, logistic regression (LR), LightGBM, random forest (RF), AdaBoost, Gaussian naive Bayes (GNB), multilayer perceptron (MLP), support vector machine (SVM), and k-nearest neighbors (KNN). The data set was partitioned into an 80% training set (n = 621) and a 20% independent testing set (n = 156). Cross-validation (CV) facilitated hyperparameter tuning and internal validation of the optimal model. Performance metrics included the area under the receiver operating characteristic curve (AUC), Brier score, accuracy, sensitivity, specificity, and F1 score. The RF model demonstrated superior performance, with AUCs of 0.992 (training) and 0.907 (validation), while the reconstructed model achieved AUCs of 0.944 (training) and 0.945 (validation), maintaining an AUC of 0.863 in the testing set. Calibration curves confirmed a strong alignment between observed and predicted probabilities. Decision curve analysis indicated that the RF model provided the highest net benefit across threshold probabilities. The SHAP algorithm identified RPR, PLT, HBV DNA, ALT, and TBA as critical predictors. This interpretable ML model enhances early HBV-LC prediction and supports clinical decision-making in resource-limited settings.

Association Between Liver Fibrosis and Risk of Incident Stroke and Mortality: A Large Prospective Cohort Study

BACKGROUND

There is a well-established relationship between liver conditions and cardiovascular diseases. However, uncertainty persists regarding the contribution of liver fibrosis to major stroke types including ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage at the population level.

METHODS

In this large prospective cohort study, participants without previous stroke or coronary heart disease at baseline from the UK Biobank were included. We identified participants at high probability of advanced liver fibrosis using the Fibrosis-4 index >2.67 or aspartate aminotransferase to platelet ratio index ≥1.0. Multivariable Cox proportional hazard regression analyses were conducted to estimate hazard ratios (HRs) for liver fibrosis with the incidence of major stroke types, stroke-related death, and all-cause death.

RESULTS

Among 379 953 participants (mean age, 56.2 [SD, 8.1] years; 44.6% men), 7396 (1.9%) had a Fibrosis-4 index >2.67 at baseline. During a median follow-up of 12.75 (interquartile range, 12.03-13.48) years, 7143 (1.9%) incident stroke cases were documented. Advanced liver fibrosis assessed by the Fibrosis-4 index was associated with an increased risk of ischemic stroke (HR, 1.94 [95% CI, 1.70-2.22]), intracerebral hemorrhage (HR, 2.14 [95% CI, 1.63-2.81]), subarachnoid hemorrhage (HR, 1.90 [95% CI, 1.27-2.84), stroke-related death (HR, 2.20 [95% CI, 1.73-2.80]), and all-cause death (HR, 2.59 [95% CI, 2.46-2.73]). Using the aspartate aminotransferase to platelet ratio index as an alternative score, liver fibrosis was correlated with magnified risk of intracerebral hemorrhage (HR, 3.76 [95% CI, 2.38-5.93]) and subarachnoid hemorrhage (HR, 3.05 [95% CI, 1.51-6.13]) compared with ischemic stroke (HR, 1.58 [95% CI, 1.17-2.14]). Restricted cubic spline analysis showed nonlinear associations of the Fibrosis-4 index and aspartate aminotransferase to platelet ratio index with stroke incidence and all-cause death.

CONCLUSIONS

Liver fibrosis is associated with increased risk of incident stroke and death among people without previous stroke or cardiovascular events, with particularly greater risk of intracerebral hemorrhage and subarachnoid hemorrhage. Noninvasive indices of liver fibrosis may serve as an easily accessible marker to detect individuals facing elevated risk of stroke and death in the primary prevention settings.

Longitudinal Evaluation of Individuals With Severe Alpha-1 Antitrypsin Deficiency (Pi∗ZZ Genotype)

BACKGROUND & AIMS

Homozygous Pi∗Z mutation in alpha-1 antitrypsin (Pi∗ZZ genotype) predisposes to pulmonary loss-of-function and hepatic gain-of-function injury. To facilitate selection into clinical trials typically targeting only 1 organ, we systematically evaluated an international, multicenter, longitudinal, Pi∗ZZ cohort to uncover natural disease course and surrogates for future liver- and lung-related endpoints.

METHODS

Cohort 1 recruited 737 Pi∗ZZ individuals from 25 different centers without known liver comorbidities who received a baseline clinical and laboratory assessment as well as liver stiffness measurement (LSM). A follow-up interview was performed after at least 6 months. Cohort 2 consisted of 135 Pi∗ZZ subjects without significant liver fibrosis, who received a standardized baseline and follow-up examination at least 2 years later, both including LSM.

RESULTS

During 2634 patient-years of follow-up, 39 individuals died, with liver and lung being responsible for 46% and 36% of deaths, respectively. Forty-one Pi∗ZZ subjects who developed a hepatic endpoint presented with significantly higher baseline liver fibrosis surrogates, that is, LSM (24 vs 5 kPa, P < .001) and aspartate aminotransferase-to-platelet ratio index (1.1 vs 0.3 units, P < .001). Liver-related endpoints within 5 years were most accurately predicted by LSM (area under the curve 0.95) followed by aspartate aminotransferase-to-platelet ratio index (0.92). Baseline lung parameters displayed only a moderate predictive utility for lung-related endpoints within 5 years (forced expiratory volume in the first second area under the curve 0.76). Fibrosis progression in those with no/mild fibrosis at baseline was rare and primarily seen in those with preexisting risk factors.

CONCLUSIONS

Noninvasive liver fibrosis surrogates accurately stratify liver-related risks in Pi∗ZZ individuals. Our findings have direct implications for routine care and future clinical trials of Pi∗ZZ patients.

Identifying and Linking Patients At Risk for MASLD with Advanced Fibrosis to Care in Primary Care

BACKGROUND AND AIMS

Severity of fibrosis is the driver of liver-related outcomes in metabolic dysfunction-associated steatotic liver disease (MASLD), and non-invasive testing such as fibrosis-4 (FIB-4) score is utilized for risk stratification. We aimed to determine if primary care patients at risk for MASLD and advanced fibrosis were evaluated with subsequent testing. A secondary aim was to determine if at-risk patients with normal aminotransferases had advanced fibrosis.

METHODS

Primary care patients at increased risk for MASLD with advanced fibrosis (n = 91,914) were identified using previously established criteria. Patients with known alternative/concomitant etiology of liver disease or cirrhosis were excluded. The study cohort included patients with calculated FIB-4 score in 2020 (n = 52,006), and stratified into low, indeterminate, and high likelihood of advanced fibrosis. Among those at indeterminate/high risk, rates of subsequent testing were measured.

RESULTS

Risk stratification with FIB-4 characterized 77% (n = 40,026) as low risk, 17% (n = 8847) as indeterminate, and 6% (n = 3133) as high risk. Among indeterminate/high-risk patients (n = 11,980), 78.7% (n = 9433) had aminotransferases within normal limits, 0.95% (n = 114) had elastography, and 8.2% (n = 984) were referred for subspecialty evaluation.

CONCLUSION

In this cohort of primary care patients at risk for MASLD with fibrosis, the FIB-4 score identified a substantial proportion of indeterminate/high-risk patients, the majority of which had normal aminotransferase levels. Low rates of subsequent testing were observed. These data suggest that a majority of patients at increased risk for liver-related outcomes remain unrecognized and highlight opportunities to facilitate their identification.

Associations between neighborhood built-environment characteristics and hepatic steatosis: The Multi-Ethnic Study of Atherosclerosis

OBJECTIVE

To characterize the spatio-temporal association between features of the built environment and subclinical liver disease.

DESIGN

We used data from a large community-based population, the Multi-Ethnic Study of Atherosclerosis (2000-2002, N = 5542) with linked historical residential data that characterized past exposure to alcohol outlets (bars and liquor stores), healthy foods stores, and physical activity facilities (1990-2001). We examined whether and how past residential relate to hepatic steatosis (proxied by liver attenuation measured using computed tomography, with lower attenuation indicating higher hepatic steatosis). Hepatic steatosis is the most common.

RESULTS

We found significant associations between past residential exposure to neighborhood alcohol outlets, healthy food and physical activity resources, and hepatic steatosis. The spatial scale where the association between these features of the built environment and hepatic steatosis operate lies within 3 km (∼2 miles). The average association on liver attenuation per additional bar, liquor, healthy food store, and physical activity facility within a 2-mile buffer, were: -0.06 (95% CI -0.09, -0.03), -0.02 (95% CI -0.04, -0.009), 0.05 (95% CI 0.02, 0.07), 0.02 (95% CI 0.01, 0.04), respectively, in the preceding year of the measurement of hepatic steatosis. Furthermore, the association and spatial scale remains consistent ten years prior to the measurement of hepatic steatosis.

CONCLUSION

Our results suggest that modifying neighborhood environments (decreasing alcohol outlets and improving access to healthy food and physical activity) may represent an effective population-wide approach to reduce liver-related morbidity.

Multisociety consensus recommendations on hepatitis delta virus infection

ABSTRACT

Hepatitis D virus (HDV) prevalence data and country-specific HDV guidelines are not widely available in the Gulf Cooperation Council (GCC) states. We developed consensus recommendations to guide healthcare professionals, policymakers, and researchers in improving HDV management and patient health outcomes in three GCC states: Kuwait, Saudi Arabia, and the United Arab Emirates. A consensus panel comprising hepatology experts (n = 6) from the three GCC societies was formed. The panel identified two broader areas related to clinical practice (screening and diagnosis, and treatment and management), addressed critical questions, and developed draft recommendations in February 2024. The strength of the final set of recommendations was subjected to consensus voting in March 2024. A majority was defined apriori with a two-thirds vote (67%). The paper outlines those recommendations alongside showcasing the current epidemiology of HDV in the GCC states, emphasizing the variability in prevalence, demographic patterns, and region-specific risk factors. It also highlights the current state of screening and diagnosis practices, identifying key obstacles, such as access to advanced screening protocols and diagnostic tools. Furthermore, HDV treatment landscape and preventative strategies are outlined, focusing on vaccination, public health initiatives, and the crucial role of public awareness and education. Ethical and sociocultural considerations are discussed, underscoring the importance of culturally sensitive healthcare practices. These recommendations present a comprehensive overview of the challenges and strategies for managing HDV in these states. Policy recommendations are provided to support HDV management, including standardizing care protocols and promoting public health measures.

Native and Gd-EOB-DTPA-Enhanced T1 mapping for Assessment of Liver Fibrosis in NAFLD: Comparative Analysis of Modified Look-Locker Inversion Recovery and Water-specific T1 mapping

RATIONALE AND OBJECTIVES

To investigate the diagnostic performance of water-specific T1 mapping for staging liver fibrosis in a non-alcoholic fatty liver disease (NAFLD) rabbit model, in comparison to Modified Look-Locker Inversion recovery (MOLLI) T1 mapping.

MATERIALS AND METHODS

60 rabbits were randomly divided into the control group (12 rabbits) and NAFLD model groups (eight rabbits per subgroup) corresponding to different durations of high-fat high cholesterol diet feeding. All rabbits underwent MRI examination including MOLLI T1 mapping and 3D multi-echo variable flip angle (VFAME- GRE) sequences were acquired before and 20 min after the administration of Gd- EOB-DTPA. Histological assessments were performed to evaluate steatosis, inflammation, ballooning, and fibrosis. Statistical analysis included the intraclass correlation coefficient, analysis of variance, spearman correlation, multiple linear regression, and receiver operating characteristic curve.

RESULTS

A moderate correlation was observed between conventional native T1 and MRI-PDFF (r = -0.513, P < 0.001), as well as between conventional native T1 and liver steatosis grades (r = -0.319, P = 0.016). However, no significant correlation was found between the native wT1 and PDFF (r = 0.137, P = 0.314), or between the native wT1 and steatosis grades (r = 0.106, P = 0.435). In the multiple regression analysis, liver fibrosis, and hepatocellular ballooning were identified as independent factors influencing native wT1 in this study (R2 =0.545, P < 0.05), while steatosis was independently associated with conventional native T1 (R2 =0.321, P < 0.05). The AUC values for native T1, native wT1, HBP T1, and HBP wT1 were 0.549(0.410-0.682), 0.811(0.684-0.903), 0.775(0.644-0.876), and 0.752(0.619-0.858) for F1 or higher, 0.581(0.441-0.711), 0.828(0.704-0.916), 0.832(0.708-0.919), and 0.854(0.734-0.934) for F2 or higher, respectively.

CONCLUSION

The native wT1 may provide a more reliable assessment of early liver fibrosis in the context of NAFLD compared to conventional native T1.

A microRNA signature for valproate-induced steatosis in human hepatocytes and its application to predict fatty liver in paediatric epileptic patients on valproate therapy

Valproate (VPA) has been the first-line, most frequently prescribed antiepileptic drug in children over the past 50 years. VPA causes, idiosyncratic hepatotoxicity in some patients, who often presents with hepatic steatosis. Experimental studies also support that VPA has high potential to induce steatosis. However, there is an apparent lack of significant hepatic problems in neuropediatric units, likely because iatrogenic liver steatosis lacks specific biomarkers. Thus, it is possible that a relevant number of children under VPA have asymptomatic fatty liver. AIMS: 1) to demonstrate VPA-induced triglyceride (TG) accumulation in cultured human upcyte hepatocytes, 2) to identify miRNAs that are deregulated by VPA and associated with TG levels in these cells, and 3) to test these miRNAs, as potential non-invasive biomarkers, in plasma of paediatric epileptic patients on VPA, to identify those with a potential risk of liver steatosis. Human upcyte hepatocytes were exposed to subcytotoxic VPA concentrations. Hepatocytes increased intracellular TGs by 27 % and 45 % after 2 and 4 mM VPA for 24 h. The profiling of cellular miRNAs by microarray analysis after 4 mM VPA identified 43 deregulated human miRNAs (fold-change > 1.5 or < -1.5; FDR p<0.05). Some of them (n=11), which were validated by RTqPCR and showed correlation (Pearson r≥ 0.6) with intracellular TG levels, were selected as potential VPA-induced steatosis biomarkers. Next, we investigated the expression of these miRNAs in human plasma and found that 9 of them could be reliably quantified by RTqPCR: miR-485-3p, miR-127-3p, miR-30a-3p, miR-92b-3p, miR-212-3p, miR-182-5p, miR-183-5p, miR-500a-5p and miR-675-5p. Screening of this 9-miRNA signature in 80 paediatric epileptic patients on VPA identified 18 patients (23 %) that clustered separately because of important alterations in the selected plasma miRNAs. These patients were younger and had higher VPA blood concentrations and serum liver enzyme levels. In conclusion, VPA induced both TG accumulation and deregulation of a set of miRNAs in cultured human hepatocytes. Nine of these miRNAs have demonstrated potential as circulating biomarkers to identify VPA-induced steatosis in epileptic patients, which should require closer clinical follow-up.

Impact of hepatic steatosis on liver stiffness measurement by vibration-controlled transient elastography and its diagnostic performance for identifying liver fibrosis in patients with chronic hepatitis B

OBJECTIVES

To explore the impact of hepatic steatosis measured by MRI-proton density fat fraction (MRI-PDFF) on liver stiffness measurement (LSM) value and its diagnostic performance for staging liver fibrosis in patients with chronic hepatitis B (CHB).

METHODS

A total of 914 patients with CHB who underwent liver biopsy and MRI-PDFF were retrospectively reviewed. The influence of MRI-PDFF on LSM value was assessed using univariate and multivariate linear analyses. To assess the influence of liver steatosis on the diagnostic performance of LSM, a series of ROC analyses were performed and compared by stratifying patients into non-steatosis (PDFF < 5%) and steatosis (PDFF ≥ 5%) groups according to MRI-PDFF values. The effects of different LSM cut-off values on the false-positive rate in the steatosis cohort were compared using McNemar's test.

RESULTS

LSM values were significantly affected by MRI-PDFF in the entire cohort (B-coefficient: 0.003, p < 0.001), F1 cohort (B-coefficient: 0.005, p < 0.001), and F2 cohort (B-coefficient: 0.003, p = 0.002). Hepatic steatosis was not observed to have a significant influence on the ROC curve of LSM for staging liver fibrosis. Compared with using the cut-off values for the CHB cohort, using relatively higher cut-off values for hepatic steatosis significantly improved the false-positive rate of LSM in the steatosis cohort.

CONCLUSION

Steatosis significantly influenced LSM, with a higher value in the early stage of liver fibrosis but did not affect the diagnostic efficiency of LSM for staging liver fibrosis. Moreover, using relatively high cut-off values significantly improved the false-positive rate of LSM in CHB patients with steatosis.

CLINICAL RELEVANCE STATEMENT

The identified correlation between MRI-PDFF and VCTE-measured LSM is not clinically relevant since the diagnostic performance of LSM in staging liver fibrosis is not affected by steatosis. A higher cut-off should be applied in CHB patients with steatosis to improve the false-positive rate.

KEY POINTS

Steatosis can affect liver stiff measurement (LSM) values in the early stage of liver fibrosis. The diagnostic performance of LSM in staging liver fibrosis is not affected by steatosis. LSM's cutoffs should be increased in patients with steatosis to improve the false-positive rate.

Regression of hepatic fibrosis after pharmacological therapy for nonalcoholic steatohepatitis

The global incidence of nonalcoholic fatty liver disease (NAFLD) is escalating considerably. NAFLD covers a range of liver conditions from simple steatosis to the more severe form known as nonalcoholic steatohepatitis, which involves chronic liver inflammation and the transformation of hepatic stellate cells into myofibroblasts that generate excess extracellular matrix, leading to fibrosis. Hepatocyte ballooning is a key catalyst for fibrosis progression, potentially advancing to cirrhosis and its decompensated state. Fibrosis is a critical prognostic factor for outcomes in patients with NAFLD; therefore, those with substantial fibrosis require timely intervention. Although liver biopsy is the most reliable method for fibrosis detection, it is associated with certain risks and limitations, particularly in routine screening. Consequently, various noninvasive diagnostic techniques have been introduced. This review examines the increasing prevalence of NAFLD, evaluates the noninvasive diagnostic techniques for fibrosis, and assesses their efficacy in staging the disease. In addition, it critically appraises current and emerging antifibrotic therapies, focusing on their mechanisms, efficacy, and potential in reversing fibrosis. This review underscores the urgent need for effective therapeutic strategies, given the dire consequences of advanced fibrosis.

Noninvasive Tests in Assessment of Patients with Alcohol-Associated Liver Disease

Alcohol-associated liver disease (ALD) remains a significant public health concern, accounting for at least half of cirrhosis cases in Europe. Historically, liver biopsy has been considered the gold standard method for both diagnosing and staging ALD. However, in the past 3 decades, there has been a growing interest in developing noninvasive biomarkers for identifying high-risk patients prone to develop liver-related complications, including elastography methods or blood-based biomarkers. This review aims to summarize currently available noninvasive testing methods that are clinically available for assessing patients with ALD, including notably steatosis and fibrosis.

An overview of risk assessment and monitoring of malignant transformation in cirrhotic nodules

Cirrhotic liver nodules can progress to hepatocellular carcinoma (HCC) through a multi-step carcinogenesis model, with dysplastic nodules being particularly high risk. Currently, monitoring the progression of non-HCC cirrhotic nodules is primarily through dynamic observation, but there is a lack of sensitive, efficient, and convenient methods. Dynamic monitoring and risk evaluation of malignant transformation are essential for timely treatment and improved patient survival rates. Routine liver biopsies are impractical for monitoring, and imaging techniques like ultrasound, computed tomography, and magnetic resonance imaging are not suitable for all patients or for accurately assessing subcentimeter nodules. Identifying serum biomarkers with high sensitivity, specificity, and stability, and developing a multi-index evaluation model, may provide a more convenient and efficient approach to monitoring pathological changes in cirrhotic nodules.

Cellular fibronectin-targeted fluorescent aptamer probes for early detection and staging of liver fibrosis

Liver fibrosis is a key process in the progression of chronic liver disease to cirrhosis. Currently, early diagnosis and precise staging of liver fibrosis remain great challenges. Extracellular matrix (ECM) molecules expressed specifically during liver fibrosis are ideal targets for bioimaging and detection of liver fibrosis. Here, we report that fluorescent probes based on a nucleic acid aptamer (ZY-1) targeting cellular fibronectin (cFN), a critical ECM molecule significantly accumulating during liver fibrosis, are promising bioimaging agents for the staging of liver fibrosis. In the work, the outstanding binding affinity of ZY-1 to cFN was validated through an in vitro model of human-derived hepatic stellate cells (HSCs). Subsequently, we constructed different ZY-1-based fluorescent probes and explored the real-time imaging performance of these fluorescent probes in CCl4-induced mouse models of different liver fibrosis stages. The ZY-1-based fluorescent probes, for the first time, effectively identified and distinguished early-stage liver fibrosis (stage 3 of Ishak 6) from advanced liver fibrosis (stage 5 of Ishak 6). The proof-of-concept study provides compelling evidences that ZY-1-based probes are a promising tool for the early diagnosis and staging of liver fibrosis and paves the way for further development of clinical-related diagnosis strategies for fibrotic diseases of the liver and other organs. STATEMENT OF SIGNIFICANCE: Currently, early diagnosis and accurate staging of liver fibrosis continue to present significant challenges. This study demonstrates that fluorescent probes based on the nucleic acid aptamer ZY-1, which targets cellular fibronectin (cFN)-a crucial extracellular matrix (ECM) molecule that significantly accumulates during liver fibrosis-are promising bioimaging agents for staging liver fibrosis. The ZY-1-based fluorescent probes effectively identified and differentiated early-stage liver fibrosis from advanced liver fibrosis. This proof-of-concept study not only provides compelling evidence that ZY-1-based probes show promise for the early diagnosis and staging of liver fibrosis but also paves the way for further investigations into the use of ZY-1 in detecting other diseases associated with cFN.

Dose-response relationship between serum N-glycan markers and liver fibrosis in chronic hepatitis B

BACKGROUND

Evaluation of liver fibrosis played a monumental role in the diagnosis and monitoring of chronic hepatitis B (CHB). We aimed to explore the value of serum N-glycan markers in liver fibrosis.

METHODS

This multi-center (33 hospitals) study recruited 760 treatment-naïve CHB patients who underwent liver biopsy. Serum N-glycan markers were analyzed by DNA sequencer-assisted fluorophore-assisted with capillary electrophoresis (DSA-FACE) technology. First, we explore the relationship between 12 serum N-glycan markers and the fibrosis stage. Then, we developed a Px score for diagnosing significant fibrosis using the LASSO regression. Next, we compared the diagnostic performances between Px, LSM, APRI, and FIB-4. Finally, we explored the relationships between glycosyltransferase gene and liver fibrosis with RNA-transcriptome sequencing.

RESULTS

We included 622 CHB participants: male-dominated (69.6%); median age 42.0 (IQR 34.0-50.0); 287 with normal ALT; 73.0% with significant fibrosis. P5(NA2), P8(NA3), and P10(NA4) were opposite to the degree of fibrosis, while other profiles (except for P0[NGA2]) increased with the degree of fibrosis. Seven profiles (P1[NGA2F], P2[NGA2FB], P3[NG1A2F], P4[NG1A2F], P7[NA2FB], P8[NA3], and P9[NA3Fb]) were selected into Px score. Px score was associated with an increased risk of significant fibrosis (for per Px score increase, the risk of significant fibrosis was increased by 3.54 times (OR = 4.54 [2.63-7.82]) in the fully-adjusted generalized linear model. p for trend was <0.001. The diagnostic performance of the Px score was superior to others. Glycosyltransferase genes were overexpressed in liver fibrosis, and glycosylation and glycosyltransferase-related pathways were significantly enriched.

CONCLUSIONS

Serum N-glycan markers were positively correlated with liver fibrosis. Px score had good performance in distinguishing significant fibrosis.

Validation study of age-independent fibrosis score (Fibrosis-3 index) in patients with metabolic dysfunction-associated steatotic liver disease

BACKGROUND AND AIMS

Because the accuracy of the Fibrosis-4 (FIB-4) index for predicting liver fibrosis changes with age, the need for different cut-offs in various age groups has frequently been discussed. We developed the age-independent score, the Fibrosis-3 (FIB-3) index, and have shown its usefulness in patients with metabolic dysfunction-associated steatotic liver disease (MASLD). This study aimed to validate the diagnostic ability of the FIB-3 index to predict fibrosis progression using a large new patient cohort.

METHODS

The ability of the FIB-3 index to predict liver fibrosis was analyzed by comparing it with that of the FIB-4 index using data from 1398 patients with MASLD enrolled in the Asia-based clinical outcome NAFLD study.

RESULTS

The areas under the receiver operating characteristic curves for predicting fibrosis stage F3 or higher were not different between the FIB-3 and FIB-4 indices in the entire cohort. Using the single ideal cut-offs of the indices (3.41 for FIB-3 index and 2.01 for FIB-4 index), the predictive accuracy of the FIB-3 index was not significantly different from that of the FIB-4 index among patients aged <60 years; however, the accuracy of the FIB-3 index was significantly higher than that of the FIB-4 index in those aged ≥60 years (0.645 and 0.529, respectively; p < 0.0001).

CONCLUSION

The high ability of the FIB-3 index with a single cut-off to predict liver fibrosis in patients with MASLD was confirmed. The FIB-3 index could serve as a useful tool for assessing liver fibrosis regardless of age.

Liver fibrosis in inflammatory arthritis patients treated with methotrexate and hydroxychloroquine: A FIB-4 index analysis

OBJECTIVES

To evaluate the risk of liver fibrosis and associated factors with the non-invasive fibrosis score-4 (FIB-4) index in patients with inflammatory arthritis using methotrexate (MTX).

METHODS

Patients with rheumatoid arthritis (RA) and psoriatic arthritis (PsA) who were followed up in the rheumatology outpatient clinic, who were on methotrexate only and for whom FIB-4 index was could be calculated at methotrexate initiation and follow-up were included. The FIB-4 index was calculated according to the following formula: age (years) × AST(IU/L)/(platelet count(10 (9)/L) × √ALT(IU/L)). The patients' demographics, comorbidities, other treatments, cumulative MTX dose, and reasons for MTX cessation were assessed. For the multivariate analysis, possible factors associated with intermediate-high risk FIB-4 index at last visit were determined.

RESULTS

A total of 107 patients were enrolled in the study, of whom 82 (76.6%) had RA and 25 (23.4%) had PsA. At the initiation of MTX, 24 (22.4%) patients had intermediate-high risk FIB-4 index. Comorbidities and the rate of ≥3-4 Charlson comorbidity index were more common in patients with intermediate-high risk FIB-4 index. A total of 37 (34.5%) patients had intermediate-high risk FIB-4 index at the last visit after median 3.6 (0.3-22.06) years follow-up. The median cumulative MTX dose was 2550 mg (1050-13.991). Cumulative MTX dose [OR 1.18 (1.01-1.33), p = .03] and diabetes mellitus [OR 4.60 (1.74-12.50), p = .002] were associated factors with intermediate-high risk FIB-4 index. The concomitant use of hydroxychloroquine (HCQ) was found to be a low-risk factor for FIB-4 index [OR 0.28 (0.10-0.78) p = .015].

CONCLUSION

The FIB-4 index is a non-invasive method that can be used in daily rheumatology practice for the evaluation and follow-up of patients who will use methotrexate. Comorbidities and cumulative MTX dose seem to be related with the risk of liver fibrosis. Concomitant use of HCQ with MTX may reduce the risk of liver fibrosis.

Comparison of the diagnostic efficacy between virtual portal pressure gradient and hepatic venous pressure gradient in patients with cirrhotic portal hypertension

OBJECTIVES

This study aimed to evaluate the performance of virtual portal pressure gradient (vPPG) and its associated hemodynamic parameters of 3-dimensional (3D) model in patients with cirrhosis.

METHODS

Seventy cirrhotic patients who underwent both hepatic venous pressure gradient (HVPG) measurement and vPPG calculation were prospectively collected. The ideal-state model (ISM; n = 44) was defined by sinusoidal PH without hepatic vein shunt or portal vein thrombosis, whereas those not conforming to the criteria were classified as non-ISM (n = 26). Correlation analyses were conducted to determine the relationship between vPPG or its associated 3D hemodynamic parameters and HVPG. The diagnostic and predictive performance of vPPG and HVPG for cirrhotic-related complications was evaluated using the receiver operating characteristic (ROC) curve and Kaplan-Meier analysis.

RESULTS

In the ISM group, vPPG-associated hemodynamic parameters including total branch cross-sectional area (S2), average branch cross-sectional area (S), and average portal vein model length (h) were correlated with HVPG (r = 0.592, 0.536, -0.497; all p < 0.001), whereas vPPG was strongly correlated with HVPG (r = 0.832, p < 0.001). In the non-ISM group, vPPG, S2, S, and h were not related to HVPG (all p > 0.05). In the ISM group, both vPPG and HVPG showed significant diagnostic and predictive capabilities for cirrhosis-related complications. While in the non-ISM group, the diagnostic accuracy and predictive efficacy of vPPG surpassed those of HVPG.

CONCLUSION

HVPG exhibited superior diagnostic and predictive efficacy for cirrhotic PH in the ISM, whereas vPPG showed enhanced performance in non-ISM.

Vibration-controlled transient elastography for significant fibrosis in treatment-naïve chronic hepatitis B patients: A systematic review and meta-analysis

BACKGROUNDS/AIMS

Accurate diagnosis of significant liver fibrosis in patients with chronic hepatitis B (CHB) is crucial when determining whether to initiate antiviral treatment (AVT). We conduct a meta-analysis to assess the diagnostic performance of vibration-controlled transient elastography (VCTE) for significant liver fibrosis in AVT-naïve CHB patients with serum alanine transaminase (ALT) levels within 5-fold the upper limit of normal (ULN).

METHODS

The Ovid-Medline, EMBASE, Cochrane, and KoreaMed databases were searched to identify studies that compared the performance of VCTE and liver biopsy (reference standard) when diagnosing significant liver fibrosis (≥F2) in AVT-naïve CHB patients with ALT within 5-fold the ULN. A hierarchical summary receiver operating characteristic curve (HSROC) and bivariate model were performed to evaluate the diagnostic performance of VCTE in the meta-analysis.

RESULTS

Eight studies (2,003 patients) were included. The summary sensitivity and specificity for diagnosis of significant liver fibrosis were 0.78 (95% confidence interval [CI], 0.66-0.86) and 0.72 (95% CI, 0.60-0.82), respectively. The HSROC for the diagnosis of significant liver fibrosis was 0.81 (95% CI, 0.72-0.86). The optimal cutoff value of VCTE for diagnosis of significant liver fibrosis was 7.7 kPa with a sensitivity of 0.64 (95% CI, 0.50-0.76) and specificity of 0.83 (95% CI, 0.72-0.90).

CONCLUSION

Our study demonstrated that VCTE has an acceptable diagnostic performance for significant liver fibrosis in AVT-naïve CHB patients with ALT within 5-fold the ULN.

Alpha-1 antitrypsin deficiency and Pi*Z allele as important co-factors in the development of liver fibrosis

BACKGROUND

Alpha-1 antitrypsin deficiency (AATD) is a codominant autosomal hereditary condition that predisposes patients to the development of lung and/or liver disease, and Pi*Z allele is the most clinically relevant mutation.

AIM

To evaluate the impact of clinical parameters and AATD phenotypes, particularly the Pi*Z allele, in liver fibrosis.

METHODS

Cross-sectional cohort study including consecutive patients with AATD followed in Pulmonology or Hepatology consultation.

RESULTS

Included 69 patients, 49.3% had Pi*MZ phenotype and 10.1% Pi*ZZ. An age ≥ 55 years, age at diagnosis ≥ 41 years and AAT at diagnosis < 77 mg/dL predicted a nonalcoholic fatty liver disease fibrosis score (NFS) not excluding advanced fibrosis [area under the curve (AUC) = 0.840, P < 0.001; AUC = 0.836, P < 0.001; AUC = 0.681, P = 0.025]. An age ≥ 50 years and age at diagnosis ≥ 41 years predicted a fibrosis-4 index of moderate to advanced fibrosis (AUC = 0.831, P < 0.001; AUC = 0.795, P < 0.001). Patients with hypertension, type 2 diabetes mellitus (DM), dyslipidaemia, metabolic syndrome, and regular alcohol consumption were more likely to have a NFS not excluding advanced fibrosis (P < 0.001, P = 0.002, P = 0.008, P < 0.001, P = 0.033). Patients with at least one Pi*Z allele and type 2 DM were 8 times more likely to have liver stiffness measurement ≥ 7.1 kPa (P = 0.040).

CONCLUSION

Risk factors for liver disease in AATD included an age ≥ 50 years, age at diagnosis ≥ 41 years, metabolic risk factors, regular alcohol consumption, at least one Pi*Z allele, and AAT value at diagnosis < 77 mg/dL. We created an algorithm for liver disease screening in AATD patients to use in primary care, selecting those to be referred to Hepatology consultation.

Pericardial Diseases and Best Practices for Pericardiectomy: JACC State-of-the-Art Review

Remarkable advances have occurred in the understanding of the pathophysiology of pericardial diseases and the role of multimodality imaging in this field. Medical therapy and surgical options for pericardial diseases have also evolved substantially. Pericardiectomy is indicated for chronic or irreversible constrictive pericarditis, refractory recurrent pericarditis despite optimal medical therapy, or partial agenesis of the pericardium with a complication (eg, herniation). A multidisciplinary evaluation before pericardiectomy is essential for optimal patient outcomes. Overall, given the good outcomes reported, radical pericardiectomy on cardiopulmonary bypass, if feasible, is the preferred approach. Due to patient complexity, as well as the technical aspects of the surgery, pericardiectomy should be performed at high-volume centers that have the required expertise. The current review highlights the essential features of this multidisciplinary approach from diagnosis to recovery in patients undergoing pericardiectomy.

Pathological biopsy strategy by regulating intracellular ROS to precisely differentiate cancer cells from diseased tissues

We have developed an innovative pathological biopsy strategy by expanding the differences of ROS levels among cancer cells, inflammatory cells and normal cells using cross-linked lipoic acid vesicles loaded with vitamin C (VC@cLAVs), combined with chemiluminescence imaging technology. By analyzing the different trends of intracellular chemiluminescence intensity, the three types of cells were quickly and accurately differentiated from diseased tissues, thus holding clinical tumor diagnostic potential.

Automated hepatic steatosis assessment on dual-energy CT-derived virtual non-contrast images through fully-automated 3D organ segmentation

PURPOSE

To evaluate the efficacy of volumetric CT attenuation-based parameters obtained through automated 3D organ segmentation on virtual non-contrast (VNC) images from dual-energy CT (DECT) for assessing hepatic steatosis.

MATERIALS AND METHODS

This retrospective study included living liver donor candidates having liver DECT and MRI-determined proton density fat fraction (PDFF) assessments. Employing a 3D deep learning algorithm, the liver and spleen were automatically segmented from VNC images (derived from contrast-enhanced DECT scans) and true non-contrast (TNC) images, respectively. Mean volumetric CT attenuation values of each segmented liver (L) and spleen (S) were measured, allowing for liver attenuation index (LAI) calculation, defined as L minus S. Agreements of VNC and TNC parameters for hepatic steatosis, i.e., L and LAI, were assessed using intraclass correlation coefficients (ICC). Correlations between VNC parameters and MRI-PDFF values were assessed using the Pearson's correlation coefficient. Their performance to identify MRI-PDFF ≥ 5% and ≥ 10% was evaluated using receiver operating characteristic (ROC) curve analysis.

RESULTS

Of 252 participants, 56 (22.2%) and 16 (6.3%) had hepatic steatosis with MRI-PDFF ≥ 5% and ≥ 10%, respectively. LVNC and LAIVNC showed excellent agreement with LTNC and LAITNC (ICC = 0.957 and 0.968) and significant correlations with MRI-PDFF values (r = - 0.585 and - 0.588, Ps < 0.001). LVNC and LAIVNC exhibited areas under the ROC curve of 0.795 and 0.806 for MRI-PDFF ≥ 5%; and 0.916 and 0.932, for MRI-PDFF ≥ 10%, respectively.

CONCLUSION

Volumetric CT attenuation-based parameters from VNC images generated by DECT, via automated 3D segmentation of the liver and spleen, have potential for opportunistic hepatic steatosis screening, as an alternative to TNC images.

Early and accurate diagnosis of steatotic liver by artificial intelligence (AI)-supported ultrasonography

OBJECTIVES

Steatotic liver disease is the most frequent chronic liver disease worldwide. Ultrasonography (US) is commonly employed for the assessment and diagnosis. Few information is available on the possible use of artificial intelligence (AI) to ameliorate the diagnostic accuracy of ultrasonography.

MATERIALS AND METHODS

An AI-based algorithm was developed using a dataset of US images. We prospectively enrolled 134 patients for algorithm validation. Patients underwent abdominal US and Proton Density Fat Fraction MRI scans (MRI-PDFF), assumed as reference technique. The hepatorenal index was manually calculated (HRIM) by 4 operators. An automatic hepatorenal index (HRIA) was obtained by the algorithm. The accuracy of HRIA to discriminate steatosis grades was evaluated by ROC analysis using MRI-PDFF cut-offs.

RESULTS

Overweight was 40 % of subjects (BMI 26.4 kg/cm2). The median HRIA was 1.11 (IQR 0.32) and the average of 4 manually calculated HRIM was 1.08 (IQR 0.26), with a 15 % inter-operator variability. Both HRIA (R = 0.79, P < 0.0001) and HRIM (R = 0.69, P < 0.0001) significantly correlated with liver fat percentage (MRI-PDFF). According to MRI-PDFF, 32 % of enrolled subjects had steatosis. Discrimination capacity by AUC between patient with steatosis and patient without steatosis was better for HRIA than HRIM (AUC: 0.87 vs. 0.82, respectively). ROC analysis showed an AUC = 0.98 for HRIA with 1.64 cut-off in distinguishing between mild and moderate/severe groups.

CONCLUSIONS

The use of AI improves accuracy and speed of ultrasonography in the diagnosis of liver steatosis. Further studies should evaluate the routine use of this technique in the management of liver steatosis at high cardio-metabolic risk.

FibroScan® versus Biochemical Scores: A Study of Liver Fibrosis in HIV with HBV Co-Infection

The study aimed to determine liver fibrosis in human immunodeficiency virus (HIV) positive individuals using transient elastography (FibroScan®), Fibrosis-4 (FIB-4) score, and aspartate aminotransferase (AST) to Platelet Ratio Index (APRI) in the HIV Department from Infectious Diseases Hospital "Victor Babeș" Craiova, Romania. Of the analyzed HIV-positive subjects (n = 161), 93 (57.76%) had HIV mono-infection, and 68 (42.24%) had Hepatitis B Virus (HBV) co-infection. The prevalence of advanced liver fibrosis was higher (F2: 11.76% and F3: 13.24%, F4: 4.41%) in the HIV-HBV co-infected group compared to the HIV mono-infected group. The univariate and multivariate analysis identified HBV co-infection (OR = 5.73) male sex (OR = 5.34), serum aspartate amino-transferase levels (Pearson's rho = 0.273), low platelet count (Pearson's rho = -0.149) and erythrocyte sedimentation rate (OR = 1.030) as risk factors for the presence of liver fibrosis. Body mass index (OR = 1.08), serum lipid levels (OR = 0.96), viral load at diagnosis (OR = 1.00005), and low CD4+ cell count (OR = 0.977) were also correlated with liver fibrosis. The FIB-4 and APRI scores were strongly correlated with each other. In conclusion, HBV co-infection seems to be a determinant factor for liver fibrosis development in people living with HIV, together with other risk factors.

Simulating federated learning for steatosis detection using ultrasound images

We aimed to implement four data partitioning strategies evaluated with four federated learning (FL) algorithms and investigate the impact of data distribution on FL model performance in detecting steatosis using B-mode US images. A private dataset (153 patients; 1530 images) and a public dataset (55 patient; 550 images) were included in this retrospective study. The datasets contained patients with metabolic dysfunction-associated fatty liver disease (MAFLD) with biopsy-proven steatosis grades and control individuals without steatosis. We employed four data partitioning strategies to simulate FL scenarios and we assessed four FL algorithms. We investigated the impact of class imbalance and the mismatch between the global and local data distributions on the learning outcome. Classification performance was assessed with area under the receiver operating characteristic curve (AUC) on a separate test set. AUCs were 0.93 (95% CI 0.92, 0.94) for source-based partitioning scenario with FedAvg, 0.90 (95% CI 0.89, 0.91) for a centralized model, and 0.83 (95% CI 0.81, 0.85) for a model trained in a single-center scenario. When data was perfectly balanced on the global level and each site had an identical data distribution, the model yielded an AUC of 0.90 (95% CI 0.88, 0.92). When each site contained data exclusively from one single class, irrespective of the global data distribution, the AUC fell in the range of 0.34-0.70. FL applied to B-mode US images provide performance comparable to a centralized model and higher than single-center scenario. Global data imbalance and local data heterogeneity influenced the learning outcome.

Alcohol use disorders and liver fibrosis: an update

Alcoholic liver disease (ALD) is currently, worldwide, the second most common cause of human fatalities every year. Alcohol use disorders (AUDs) lead to 80% of hepatotoxic deaths, and about 40% of cases of cirrhosis are alcohol-related. An acceptable daily intake (ADI) of ethanol is hard to establish and studies somewhat controversially recommend a variety of dosages of ADI, whilst others regard any intake as dangerous. Steatohepatitis should be viewed as "the rate limiting step": generally, it can be overcome by abstinence, although in some patients, abstinence has little effect, with the risk of fibrosis, leading in some cases to hepatocellular carcinoma (HCC). Chronic alcoholism can also cause hypercortisolism, specifically pseudo-Cushing Syndrome, whose diagnosis is challenging. If fibrosis is spotted early, patients may be enrolled in detoxification programs to achieve abstinence. Treatment drugs include silybin, metadoxine and adenosyl methionine. Nutrition and the proper use of micronutrients are important, albeit often overlooked in ALD treatment. Other drugs, with promising antifibrotic effects, are now being studied. This review deals with the clinical and pathogenetic aspects of alcohol-related liver fibrosis and suggests possible future strategies to prevent cirrhosis.

Combining T1rho and advanced diffusion MRI for noninvasively staging liver fibrosis: an experimental study in rats

PURPOSE

To investigate the value of imaging parameters derived from T1 relaxation times in the rotating frame (T1ρ or T1rho), diffusion kurtosis imaging (DKI) and intravoxel incoherent motion (IVIM) in assessment of liver fibrosis in rats and propose an optimal diagnostic model based on multiparametric MRI.

METHODS

Thirty rats were divided into one control group and four fibrosis experimental groups (n = 6 for each group). Liver fibrosis was induced by administering thioacetamide (TAA) for 2, 4, 6, and 8 weeks. T1ρ, mean kurtosis (MK), mean diffusivity (MD), perfusion fraction (f), true diffusion coefficient (D), and pseudo-diffusion coefficient (D*) were measured and compared among different fibrosis stages. An optimal diagnostic model was established and the diagnostic efficiency was evaluated by receiver operating characteristic (ROC) curve analysis.

RESULTS

The mean AUC values, sensitivity, and specificity of T1ρ and MD derived from DKI across all liver fibrosis stages were comparable but much higher than those of other imaging parameters (0.954, 92.46, 91.85 for T1ρ; 0.949, 92.52, 91.24 for MD). The model combining T1ρ and MD exhibited better diagnostic performance with higher AUC values than any individual method for staging liver fibrosis (≥ F1: 1.000 (0.884-1.000); ≥ F2: 0.935 (0.782-0.992); ≥ F3: 0.982 (0.852-1.000); F4: 0.986 (0.859-1.000)).

CONCLUSION

Among the evaluated imaging parameters, T1ρ and MD were superior for differentiating varying liver fibrosis stages. The model combining T1ρ and MD was promising to be a credible diagnostic biomarker to detect and accurately stage liver fibrosis.

Diagnostic performance of non-invasive liver fibrosis scores in patients with early-intermediate hepatocellular carcinoma

PURPOSE

Hepatocellular carcinoma (HCC) arises in individuals with underlying liver disease. Diagnosing the degree of hepatic fibrosis helps to determine the severity of the underlying liver disease and may influence therapeutic decisions in HCC patients. Non-invasive fibrosis scores can be used to estimate the degree of fibrosis in liver disease patients, but most of these scores were developed in patients with viral hepatitis and without HCC. This study explored the ability of the Fibrosis-4 Index (FIB-4), the AST/Platelet Ratio Index (APRI), and the AST/ALT ratio to diagnose or exclude advanced fibrosis (METAVIR F3/4 versus F0-2) in patients with early-intermediate, potentially resectable HCC.

METHODS

We retrospectively reviewed 119 patients who underwent hepatic resection for HCC at a tertiary centre (2007-2019), 75 of whom had advanced fibrosis (prevalence 63%). Histological assessment of the surgical liver specimen was used as a reference standard for the degree of fibrosis.

RESULTS

Overall diagnostic performance was highest for the FIB-4 Index, with an area under the receiver operating characteristic curve (AUROC) of 0.82, compared with 0.78 for APRI, and 0.56 for the AST/ALT ratio. Using established cut-off values, FIB-4 achieved a 90% positive predictive value at the higher cut-off (3.25) and a 90% negative predictive value at the lower cut-off (1.45).

CONCLUSION

The FIB-4 Index could reliably diagnose or exclude advanced fibrosis in patients with early-intermediate HCC, and may thus have a role in guiding therapeutic decisions in these patients.

Magnetic resonance elastography for the prediction of hepatocellular carcinoma in chronic hepatitis B

Background and Aim

Magnetic resonance elastography (MRE) is used for the evaluation of liver fibrosis; however, it remains unclear whether MRE-based liver stiffness is associated with hepatocellular carcinoma (HCC) development, particularly in patients with chronic hepatitis B.

Methods

A total of 504 patients with chronic hepatitis B receiving MRE were enrolled. The primary endpoint was the association between MRE-based liver stiffness and HCC.

Results

In a cross-sectional analysis at the time of MRE measurement, the median (interquartile range) liver stiffness values in patients with presence or history of HCC and those without HCC were 3.68 (2.89-4.96) and 2.60 (2.22-3.45) kPa, respectively, and liver stiffness was significantly higher in patients with presence or history of HCC than in those without HCC (P < 0.001). In a longitudinal analysis of patients without HCC, the 1-, 3-, and 5-year cumulative incidence of HCC in patients with liver stiffness ≥3.6 kPa and those with liver stiffness <3.6 kPa were 3.8%, 7.0%, and 22.9%, and 0%, 0.9%, and 1.5%, respectively (P < 0.001). In the multivariable analysis, MRE-based liver stiffness (per 1 kPa) or liver stiffness ≥3.6 kPa was an independent factor for HCC development with an adjusted hazard ratio (aHR) of 1.61 (95% confidence interval [CI], 1.3-2.0) or aHR of 8.22 (95% CI, 2.1-31).

Conclusion

MRE-based liver stiffness is associated with HCC risk in patients with chronic hepatitis B and may be used for the early prediction of HCC development and determination of indications for treatment.

Liver stiffness and associated risk factors among people with a history of injecting drugs: a prospective cohort study

BACKGROUND

Persons with opioid use disorders (OUD) and persons with substance use disorders (SUD) who inject substances have a reduced life expectancy of up to 25 years compared with the general population. Chronic liver diseases are a substantial cause of this. Screening strategies based on liver stiffness measurements (LSM) may facilitate early detection, timely intervention, and treatment of liver disease. This study aims to investigate the extent of chronic liver disease measured with transient elastography and the association between LSM and various risk factors, including substance use patterns, hepatitis C virus (HCV) infection, alcohol use, body mass index, age, type 2 diabetes mellitus, and high-density lipoprotein (HDL) cholesterol among people with OUD or with SUD who inject substances.

METHODS

Data was collected from May 2017 to March 2022 in a cohort of 676 persons from Western Norway. The cohort was recruited from two populations: Persons receiving opioid agonist therapy (OAT) (81% of the sample) or persons with SUD injecting substances but not receiving OAT. All participants were assessed at least once with transient elastography. A linear mixed model was performed to assess the impact of risk factors such as HCV infection, alcohol use, lifestyle-associated factors, and substance use on liver stiffness at baseline and over time. Baseline was defined as the time of the first liver stiffness measurement. The results are presented as coefficients (in kilopascal (kPa)) with 95% confidence intervals (CI).

RESULTS

At baseline, 12% (n = 83) of the study sample had LSM suggestive of advanced chronic liver disease (LSM ≥ 10 kPa). Advanced age (1.0 kPa per 10 years increments, 95% CI: 0.68;1.3), at least weekly alcohol use (1.3, 0.47;2.1), HCV infection (1.2, 0.55;1.9), low HDL cholesterol level (1.4, 0.64;2.2), and higher body mass index (0.25 per increasing unit, 0.17;0.32) were all significantly associated with higher LSM at baseline. Compared with persistent chronic HCV infection, a resolved HCV infection predicted a yearly reduction of LSM (-0.73, -1.3;-0.21) from baseline to the following liver stiffness measurement.

CONCLUSIONS

More than one-tenth of the participants in this study had LSM suggestive of advanced chronic liver disease. It underscores the need for addressing HCV infection and reducing lifestyle-related liver risk factors, such as metabolic health factors and alcohol consumption, to prevent the advancement of liver fibrosis or cirrhosis in this particular population.

Serum/Plasma Proteome in Non-Malignant Liver Disease

The liver is the central metabolic organ and produces 85-90% of the proteins found in plasma. Accordingly, the plasma proteome is an attractive source of liver disease biomarkers that reflects the different cell types present in this organ, as well as the processes such as responses to acute and chronic injury or the formation of an extracellular matrix. In the first part, we summarize the biomarkers routinely used in clinical evaluations and their biological relevance in the different stages of non-malignant liver disease. Later, we describe the current proteomic approaches, including mass spectrometry and affinity-based techniques, that allow a more comprehensive assessment of the liver function but also require complex data processing. The many approaches of analysis and interpretation and their potential caveats are delineated. While these advances hold the promise to transform our understanding of liver diseases and support the development and validation of new liver-related drugs, an interdisciplinary collaboration is needed.

Prospective comparison of an 18-gauge versus 16-gauge needle for percutaneous liver core-needle biopsy in children

PURPOSE

The objective of this study was to analyzed the impact of needle gauge (G) on the adequacy of specimens and hemorrhagic complications in pediatric patients undergoing ultrasound (US)-guided transplanted liver biopsies.

METHODS

The study included 300 consecutive biopsies performed in 282 pediatric patients (mean age 6.75 ± 3.82 years, range 0.84-17.90) between December 2020 and April 2022. All pediatric patients that referred to our institution for US-guided core-needle liver biopsy (CNLB) were randomized to undergo 16-G or 18-G CNLB. Hemorrhagic complications were qualitatively evaluated. The number of complete portal tracts (CPTs) per specimen was counted and specimen adequacy was assessed based on the American Association for the Study of Liver Diseases guidelines.

RESULTS

The incidence of bleeding was 7.00% (n = 21) and adequate specimens for accurate pathological diagnosis were obtained from 98.33% (n = 295) of patients. There was no significant difference in the incidence or amount of bleeding between the 16-G and 18-G groups (11 vs 10, p = 0.821; 35.0 mL vs 31.3 mL, p = 0.705). Although biopsies obtained using a 16-G needle contained more complete portal tracts than those obtained using an 18-G needle (20.0 vs 18.0, p = 0.029), there was no significant difference in specimen inadequacy according to needle gauge (2 vs 3, p = 1.000).

CONCLUSIONS

Biopsy with a 16-G needle was associated with a greater number of CPTs but did not increase the adequate specimen rate. There was no significant difference in the complication rate between 16-G biopsy and 18-G biopsy.

Management of alcohol withdrawal syndrome in patients with alcohol-associated liver disease

Alcohol-associated liver disease is a common and severe sequela of excessive alcohol use; effective treatment requires attention to both liver disease and underlying alcohol use disorder (AUD). Alcohol withdrawal syndrome (AWS) can be dangerous, is a common barrier to AUD recovery, and may complicate inpatient admissions for liver-related complications. Hepatologists can address these comorbid conditions by learning to accurately stage alcohol-associated liver disease, identify AUD using standardized screening tools (eg, Alcohol Use Disorder Identification Test), and assess risk for and symptoms of AWS. Depending on the severity, alcohol withdrawal often merits admission to a monitored setting, where symptom-triggered administration of benzodiazepines based on standardized scoring protocols is often the most effective approach to management. For patients with severe liver disease, selection of benzodiazepines with less dependence on hepatic metabolism (eg, lorazepam) is advisable. Severe alcohol withdrawal often requires a "front-loaded" approach with higher dosing, as well as intensive monitoring. Distinguishing between alcohol withdrawal delirium and HE is important, though it can be difficult, and can be guided by differentiating clinical characteristics, including time to onset and activity level. There is little data on the use of adjuvant medications, including anticonvulsants, dexmedetomidine, or propofol, in this patient population. Beyond the treatment of AWS, inpatient admission and outpatient hepatology visits offer opportunities to engage in planning for ongoing management of AUD, including initiation of medications for AUD and referral to additional recovery supports. Hepatologists trained to identify AUD, alcohol-associated liver disease, and risk for AWS can proactively address these issues, ensuring that patients' AWS is managed safely and effectively and supporting planning for long-term recovery.

Cognitive impairment and liver fibrosis in non-alcoholic fatty liver disease

Background

Data regarding the prevalence and phenotype of cognitive impairment in non-alcoholic fatty liver disease (NAFLD) are limited.

Objective

We assessed the prevalence and nature of cognitive deficits in people with NAFLD and assessed whether liver fibrosis, an important determinant of outcomes in NAFLD, is associated with worse cognitive performance.

Methods

We performed a prospective cross-sectional study. Patients with NAFLD underwent liver fibrosis assessment with transient elastography and the following assessments: Cognitive Change Index, Eight-Item Informant Interview to Differentiate Aging and Dementia Questionnaire (AD8), Montreal Cognitive Assessment (MoCA), EncephalApp minimal hepatic encephalopathy test and a limited National Institutes of Health Toolbox battery (Flanker Inhibitory Control and Attention Test, Pattern Comparison Test and Auditory Verbal Learning Test). We used multiple linear regression models to examine the association between liver fibrosis and cognitive measures while adjusting for relevant covariates.

Results

We included 69 participants with mean age 50.4 years (SD 14.4); 62% were women. The median liver stiffness was 5.0 kilopascals (IQR 4.0-6.9), and 25% had liver fibrosis (≥7.0 kilopascals). Cognitive deficits were common in people with NAFLD; 41% had subjective cognitive impairment, 13% had an AD8 >2, 32% had MoCA <26 and 12% had encephalopathy detected on the EncephalApp test. In adjusted models, people with liver fibrosis had modestly worse performance only on the Flanker Inhibitory Control and Attention Task (β=-0.3; 95% CI -0.6 to -0.1).

Conclusion

Cognitive deficits are common in people with NAFLD, among whom liver fibrosis was modestly associated with worse inhibitory control and attention.

Utility and impact of magnetic resonance elastography in the clinical course and management of chronic liver disease

We aimed to characterize scenarios where magnetic resonance elastography (MRE) of the liver was ordered and its impact on clinical course and management. 96 consecutive MRE examinations and subsequent encounters over 14 months were reviewed. Indication for MRE of the liver and subsequent management were abstracted from the medical record. In all cases, non-invasive assessment of liver fibrosis was the primary indication and at least one additional rationale was noted. There was a significant decrease in recommendations to undergo liver biopsy after MRE. Additionally, a greater percentage of those recommended to undergo biopsy completed the procedure after discussion of the results. Given the significant cost and rare but serious risks of liver biopsy, MRE of the liver provides an attractive, safer alternative that may have a comparable impact on management, or select cases where biopsy is essential to guide management. We demonstrate the versatility of MRE in real-world hepatology practice, including its utility as a non-invasive surrogate for liver biopsy.

Metabolic Dysfunction-Associated Fatty Liver Disease and Chronic Viral Hepatitis: The Interlink

Metabolic dysfunction-associated fatty liver disease (MAFLD) has now affected nearly one-third of the global population and has become the number one cause of chronic liver disease in the world because of the obesity pandemic. Chronic hepatitis resulting from hepatitis B virus (HBV) and hepatitis C virus (HCV) remain significant challenges to liver health even in the 21st century. The co-existence of MAFLD and chronic viral hepatitis can markedly alter the disease course of individual diseases and can complicate the management of each of these disorders. A thorough understanding of the pathobiological interactions between MAFLD and these two chronic viral infections is crucial for appropriately managing these patients. In this comprehensive clinical review, we discuss the various mechanisms of chronic viral hepatitis-mediated metabolic dysfunction and the impact of MAFLD on the progression of liver disease.

Macrophage activation syndrome with acute hepatitis in a patient with adult-onset immunodeficiency with anti-interferon gamma antibodies: a case report

BACKGROUND

Macrophage activation syndrome is a rare disorder leading to unregulated immune activity manifesting with nonspecific constitutional symptoms, laboratory abnormalities, and multiorgan involvement. We report the case of a patient who presented with acute hepatitis secondary to macrophage activation syndrome diagnosed by liver biopsy and successfully treated with intravenous immune globulin, anakinra, and rituximab.

CASE PRESENTATION

A 42-year-old Laotian woman with adult-onset immunodeficiency with anti-interferon gamma antibodies presented with a fever, headache, generalized myalgia, dark urine, and reduced appetite in the setting of family members at home with similar symptoms. Her laboratory workup was notable for evidence of acute hepatitis without acute liver failure. After an unrevealing comprehensive infectious and noninvasive rheumatologic workup was completed, a liver biopsy was performed ultimately revealing the diagnosis of macrophage activation syndrome. She was successfully treated with intravenous immune globulin, anakinra, and rituximab.

CONCLUSION

This case highlights the importance of maintaining macrophage activation syndrome on the differential of a patient with acute hepatitis of unknown etiology in the correct clinical context and the value of a liver biopsy in making a diagnosis when noninvasive testing is unrevealing.

Liver fibrosis classification from ultrasound using machine learning: a systematic literature review

PURPOSE

Liver biopsy was considered the gold standard for diagnosing liver fibrosis; however, with advancements in medical technology and increasing awareness of potential complications, the reliance on liver biopsy has diminished. Ultrasound is gaining popularity due to its wider availability and cost-effectiveness. This study examined the machine learning / deep learning (ML/DL) models for non-invasive liver fibrosis classification from ultrasound.

METHODS

Following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) protocol, we searched five academic databases using the query. We defined population, intervention, comparison, outcomes, and study design (PICOS) framework for the inclusion. Furthermore, Joana Briggs Institute (JBI) checklist for analytical cross-sectional studies is used for quality assessment.

RESULTS

Among the 188 screened studies, 17 studies are selected. The methods are categorized as off-the-shelf (OTS), attention, generative, and ensemble classifiers. Most studies used OTS classifiers that combined pre-trained ML/DL methods with radiomics features to determine fibrosis staging. Although machine learning shows potential for fibrosis classification, there are limited external comparisons of interventions and prospective clinical trials, which limits their applicability.

CONCLUSION

With the recent success of ML/DL toward biomedical image analysis, automated solutions using ultrasound are developed for predicting liver diseases. However, their applicability is bounded by the limited and imbalanced retrospective studies having high heterogeneity. This challenge could be addressed by generating a standard protocol for study design by selecting appropriate population, interventions, outcomes, and comparison.

Image-based AI diagnostic performance for fatty liver: a systematic review and meta-analysis

BACKGROUND

The gold standard to diagnose fatty liver is pathology. Recently, image-based artificial intelligence (AI) has been found to have high diagnostic performance. We systematically reviewed studies of image-based AI in the diagnosis of fatty liver.

METHODS

We searched the Cochrane Library, Pubmed, Embase and assessed the quality of included studies by QUADAS-AI. The pooled sensitivity, specificity, negative likelihood ratio (NLR), positive likelihood ratio (PLR), and diagnostic odds ratio (DOR) were calculated using a random effects model. Summary receiver operating characteristic curves (SROC) were generated to identify the diagnostic accuracy of AI models.

RESULTS

15 studies were selected in our meta-analysis. Pooled sensitivity and specificity were 92% (95% CI: 90-93%) and 94% (95% CI: 93-96%), PLR and NLR were 12.67 (95% CI: 7.65-20.98) and 0.09 (95% CI: 0.06-0.13), DOR was 182.36 (95% CI: 94.85-350.61). After subgroup analysis by AI algorithm (conventional machine learning/deep learning), region, reference (US, MRI or pathology), imaging techniques (MRI or US) and transfer learning, the model also demonstrated acceptable diagnostic efficacy.

CONCLUSION

AI has satisfactory performance in the diagnosis of fatty liver by medical imaging. The integration of AI into imaging devices may produce effective diagnostic tools, but more high-quality studies are needed for further evaluation.