In this installment of Expert Picks from EASL 2025, Ashwani Singal, MD, the medical director of liver transplantation at the University of Louisville’s Trager Transplant Center and a professor of medicine at the University of Louisville School of Medicine, in Kentucky, discusses five of his favorite abstracts from the meeting.
Abstract OS-034. Corticosteroids are ineffective in individuals with severe alcohol-associated hepatitis and early spontaneous improvement: a multicenter randomized controlled trial (Moreno et al)
Investigators conducted a randomized controlled trial at 10 hospitals in Belgium to evaluate the effect of corticosteroid treatment on mortality in patients with severe alcohol-associated hepatitis (AH) who had spontaneous improvement after hospital admission.
Adult hospitalized patients were eligible if they had severe AH (Maddrey discriminant function [mDF], =32), as indicated by biopsy, and recent-onset jaundice with spontaneous improvement, defined as more than a 10% decrease in serum bilirubin within five to 10 days after admission.
Participants were randomized to either 32 mg of methylprednisolone per day (n=38) or placebo (n=31) for 28 days. The primary end point was three-month mortality, with secondary end points of one-month mortality and infection rate.
At baseline, patients in the corticosteroid and placebo groups had similar age (52 vs. 51 years), sex distribution (76% vs. 63% male), total bilirubin (7.4 vs. 9.3 mg/dL), international normalized ratio (INR) (1.64 vs. 1.70), mDF (44 vs. 45), Model for End-Stage Liver Disease (MELD) score (21 vs. 21), and decrease in bilirubin from admission (31% vs. 27%).
At three months, the probability of survival was similar in the corticosteroid and placebo groups (83% vs. 82%). Similarly, the probability of survival at one month was comparable between the treatment groups (95% vs. 94%).
Among those in the corticosteroid group, 48% had an infection during the study period, compared with 36% in the placebo group (P=0.41).
The investigators noted that despite the study being “prematurely interrupted due to a low recruitment rate,” the results indicate that “it is unlikely that corticosteroids provide a survival benefit when bilirubin level spontaneously decreases by at least 10%.” Thus, “waiting five days after admission before deciding to start steroids seems to be a reasonable strategy.”
Dr. Singal: When patients are hospitalized with severe AH, residents, fellows, and hospitalists often think about giving steroids as the first thing and as an urgent decision. It should be recognized that corticosteroids are recommended, but they are not perfect drugs. They have side effects, so take your time to screen your population, make sure they have no infection, get cultures, give it time.
In most randomized trials, the median time to starting steroids has been four to five days because it takes time for the cultures to come back. And in that time, the body’s regenerative and hepatic immune system can result in spontaneous improvement. In that case, there may be no need to start the treatment, and you can avoid making the patient more susceptible to infections.
This is an important study because it is randomized and confirms the practice of avoiding steroid use in patients with severe AH who have spontaneous improvement of liver disease and serum bilirubin.
Abstract OS-043-YI. Clinical impact of non selective beta-blockers in patients with cirrhosis and acute kidney injury: a post hoc analysis of the international club of ascites GLOBAL-AKI study (Incicco et al)
In a post hoc analysis of the GLOBAL-AKI study, investigators evaluated the effect of treatment with nonselective beta-blockers (NSBBs) on the outcomes of patients with acute kidney injury (AKI).
The prospective GLOBAL-AKI study included patients who were hospitalized between July 2022 and May 2023 with acute decompensation of cirrhosis. Outcomes were compared between AKI patients who were taking NSBBs at diagnosis and those who were not. In addition, where data were available, outcomes were compared between those who continued NSBBs during hospitalization and those who tapered or discontinued these agents.
The investigators used propensity score matching to balance the NSBB and no-NSBB groups by age, sex, sodium-adjusted model for end-stage liver disease (MELD-Na) score, mean arterial pressure (MAP), ascites, hepatic encephalopathy, AKI stage, acute-on-chronic liver failure (ACLF) grade at AKI diagnosis, and whether patients were from low- or middle-income countries.
Among 1,238 patients with AKI, 503 were taking NSBBs at diagnosis (55% propranolol, 45% carvedilol). These patients had lower odds of AKI non-resolution (odds ratio [OR], 0.65; P=0.001) and lower hazard of 28-day mortality (subdistribution hazard ratio [sHR], 0.68; P=0.005) than patients who were not taking NSBBs at AKI diagnosis.
When assessing the association of continuing (n=123) versus tapering or withdrawing from NSBBs (n=350) during hospitalization, the investigators found that there was no statistically significant difference in the odds of AKI non-resolution (OR, 1.00; P=0.999) or the hazard of 28-day mortality (sHR, 0.56; P=0.190) between groups.
These results were consistent in subgroup analyses of patients with MAP under 70 mm Hg, serum Na under 130 mmol/L, and ACLF.
Based on these results, the investigators concluded, “Continuation of treatment with NSBBs in patients with cirrhosis and AKI does not seem to be harmful.”
Dr. Singal: This is an important study for two reasons. First, patients with cirrhosis who are on NSBBs have better outcomes when hospitalized. Second, if NSBBs are continued during the hospitalization in patients with AKI, the outcomes are unaffected. Hence, patients can continue taking these NSBBs without negatively affecting their AKI outcomes while they’re hospitalized.
These results can reassure community gastroenterologists and hepatologists that the hospitalized AKI patients they see are not more likely to have adverse outcomes because of continued NSBBs.
Abstract OS-092-YI. The transcriptomic profile of antibody-mediated rejection is independent from the presence of donor specific antibodies (Engel et al)
A multicenter team of researchers assessed the transcriptomic signature of liver biopsies in patients who underwent liver transplantation (LT) to look for differences in gene expression in cases of chronic antibody mediated rejection (cABMR), cABMR with absence of donor-specific antibodies (cABMR-DSAneg), and no histologic rejection (NHR).
Liver biopsies performed from 2018 to 2022 at the Hannover Medical School, in Germany, were evaluated for inclusion. Of 21 biopsies meeting cABMR criteria, per Banff 2016 criteria, and with enough tissue for RNA extraction, 15 were cABMR-DSAneg. In addition, 20 biopsies with NHR were used as controls.
Patients with cABMR (n=6) and cABMR-DSAneg (n=15) had similar age at LT (47 vs. 40 years; P=0.88) and time since LT (149 vs. 158 months; P=0.88), with most cases found in surveillance biopsies, as opposed to indication biopsies (cABMR, 50%; cABMR-DSAneg, 67%).
“The transcriptomic profile or cABMR was characterized by fibrogenesis, complement activation and TNF pathways,” the investigators concluded. They did not find any differences in gene expression in cABMR versus cABMR-DSAneg, “with no up- or down-regulated genes that differ between the groups.”
These results indicate that “the presence of DSA, as described in other solid organ transplants, is not necessary for the diagnosis of cABMR.”
Dr. Singal: I think this is a very important study, because we see it all the time that just the presence of DSAs in the serum in a transplant recipient triggers plasmapheresis and treatment of antibody-mediated rejection. But this is not always appropriate because, first, antibody-mediated rejection in LT is a very rare event, and second, you have to take your time to make the diagnosis. Obviously, DSAs and the titers are important, but also you should have a biopsy and C4d staining and make sure that you have proper histology, so you’re not treating somebody for antibody-mediated rejection just because they have DSAs. This study makes that point and can remind gastroenterologists that here is a proper set of findings you need to see on biopsy to make the diagnosis of antibody-mediated rejection. It’s supported by DSAs in the serum, but that alone should not be enough to make that diagnosis.
Abstract OS-097. Evaluating noninvasive diagnostic pathways for identifying MASH patients eligible for resmetirom therapy: a real-world cohort analysis (Dunn et al)
A team of researchers at six U.S. tertiary hepatology clinics compared the performance of several noninvasive tests (NITs) for diagnosis of metabolic dysfunction–associated steatohepatitis (MASH) with significant (F2) to advanced (F3) fibrosis and consideration for treatment with resmetirom (Rezdiffra, Madrigal). The investigators noted that such NITs can be used in primary care to “determine who needs referral to hepatology care.”
Patients were included in this retrospective analysis if they were treated for metabolic dysfunction–associated steatotic liver disease (MASLD) with resmetirom between March and November 2024 at one of the six study centers and met at least one of the AASLD treatment consideration criteria (liver stiffness measurement [LSM] of 8-20 kPa, Enhanced Liver Fibrosis [ELF] test result of 9.2-11.3, or liver biopsy fibrosis stage 2 or 3).
In total, 424 patients were evaluated for inclusion. Among 88.9% who underwent transient elastography, 69.8% met the AASLD LSM-based criteria. Among 13.2% who underwent ELF testing, 73.2% met the AASLD ELF-based criteria. Among 20.8% who underwent liver biopsy, 45.5% had F2 fibrosis and 46.6% had F3 fibrosis. Thus, 337 of the patients met AASLD criteria and were included in the analysis evaluating the NITs (FIB4, SAFE score, LiverRisk Score, and ALADDIN-F2-Lab score).
The proportion of the 337 patients who met AASLD criteria but were characterized as low risk by the NITs ranged from 12.5% for ALADDIN-F2-Lab, 13.0% for SAFE, 45.8% for FIB-4, and 80.2% for LiverRisk Score.
The investigators highlighted that “this real-world analysis highlights the limitations of current AASLD and EASL guidelines, which rely on FIB-4, in identifying MASH patients with significant fibrosis,” based on “nearly half” of eligible patients being characterized as low risk based on FIB-4. “In contrast,” the investigators noted, “novel scoring systems like SAFE and ALADDIN-F2-Lab, leveraging routine laboratory data, identified the majority of patients as intermediate or high risk.”
Dr. Singal: This study came up with a new machine learning model called ALLADIN-F2-Lab (Am J Gastroenterol 2025 Mar 27. doi:10.14309/ajg.0000000000003432), developed under the leadership of Dr. Winston Dunn, a transplant hepatologist at the University of Kansas Health System, which was shown to be better than the two-step strategy of FIB-4 followed by vibration-controlled transient elastography (VCTE) (FibroScan, EchoSens) for identifying F2/F3 patients. I think an important advantage of this model is that does not take VCTE into consideration, so community gastroenterologists and hepatologists who do not have access to VCTE can use the ALADDIN model and be confident in picking F2 or F3 patients who may be candidates for resmetirom treatment for MASH. They can go to a website to get the probability of F2 or higher fibrosis by plugging in values for certain variables, including VCTE (aihepatology.shinyapps.io/ALADDIN1/) or excluding VCTE (aihepatology.shinyapps.io/ALADDIN2/).
Abstract SAT406. Role of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) in mortality reduction among patients with obesity and metabolic dysfunction-associated steatotic liver disease (MASLD) cirrhosis: a global retrospective analysis (Aitharaju et al)
Using data from TriNetX, researchers assessed the association of glucagon-like peptide-1 receptor agonist (GLP-1) use with mortality and hepatocellular carcinoma (HCC) incidence in a cohort of patients with MASLD complicated by cirrhosis.
Data from 2004 to 2024 were examined, with patients split into those who were prescribed GLP-1s after diagnosis (users) and those who had never been prescribed a GLP-1 (nonusers). Patients who had cirrhosis from etiologies other than MASLD were excluded, as were patients with decompensated cirrhosis.
Among 40,646 eligible patients, 4,328 were GLP-1 users. Propensity score matching resulted in inclusion of 3,846 users and 3,846 nonusers in analyses.
The investigators found that GLP-1 users had a decreased hazard of five-year mortality (HR, 0.37; 95% CI, 0.27-0.50) and five-year incidence of HCC (HR, 0.57; 95% CI, 0.36-0.89), compared with nonusers.
Secondary outcomes were also improved in GLP-1 users, including lower incidence of liver transplant (hazard ratio [HR], 0.45; 95% CI, 0.24-0.82), sepsis (HR, 0.52; 95% CI, 0.36-0.76), inpatient admissions (HR, 0.54; 95% CI, 0.47-0.62), obesity-associated malignancies (HR, 0.64; 95% CI, 0.54-0.76), and major adverse cardiovascular events (HR, 0.84; 95% CI, 0.72-0.98).
Dr. Singal: This is an important study, as data are emerging that GLP-1s, at least in compensated cirrhosis patients, are safe. So, I think this is another example of reassuring data on GLP-1 safety and efficacy in a cirrhosis population. We know that 40% of weight loss that comes from GLP-1s is lean mass and not fat mass. But cirrhosis patients are prone to lean body mass loss due to nutritional deficiency and catabolic state. It would be relevant to have a study of GLP-1 agonists in patients with cirrhosis assessing muscle density and sarcopenia.
—Compiled and written by Natasha Albaneze, MPH
Dr. Singal is a member of the Gastroenterology & Endoscopy News editorial board.
This article is from the December 2025 print issue.