Back

A liver function test score identifies high-risk MASLD patients based on the pattern of liver enzymes

Hajaj, E.; Glusman Bendersky, A.; Braun, M.; Shlomai, A.

2024-10-27 gastroenterology
10.1101/2024.10.26.24316188 medRxiv
Show abstract

Background & AimsA cholestatic pattern of liver enzymes is associated with progressive liver disease and major adverse liver-related outcomes (MALO) among patients with Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD). We aimed to authenticate the efficacy of a newly formulated liver function test (LFT) score for distinguishing patients with cholestatic vs. hepatocellular patterns and to evaluate its prognostic utility in MASLD patients. MethodsA retrospective longitudinal study on a dataset of over 250,000 individuals diagnosed with MASLD and/or obesity with cardiovascular risk factors. Patients were categorized into cholestatic (C), mixed (M), or hepatocellular (H) patterns according to the LFT score, or the well-known R score. Long-term MALO, major adverse cardiovascular events (MACE), and all-cause mortality were tracked. ResultsThe LFT score excelled in differentiating patients into C, M, or H groups accurately. While about two-thirds of our cohort initially showed a low FIB4 (<1.3), patients in the C category experienced a higher incidence of MALO and MACE compared to those in the H category (0.5% vs. 0.2% and 7.1% vs. 3.6%, respectively) over the span of 10 years post-diagnosis. Additionally, the 15-year overall survival rate was notably lower for C patients compared to their H counterparts (63% vs. 77%, p<0.0001). The LFT score was more effective than the R score in distinguishing between H and C patients for prognostic purposes, and a baseline cholestatic pattern indicates poorer outcomes regardless of subsequent LFT changes. ConclusionsThe LFT score accurately categorizes cholestatic MASLD patients and may serve as a useful prognostic tool.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.