脂肪性肝疾患における表現型特異的炎症プロファイル：線維症同定への示唆

Liver fibrosis is a key prognostic factor in steatotic liver diseases (SLD), including metabolic dysfunction-associated steatotic liver disease (MASLD), metabolic and alcohol-associated liver disease (MetALD), and alcohol-related liver disease (ALD). This study aimed to (1) compare systemic inflammation and fibrosis across phenotypes, (2) study correlations between inflammatory markers and occurrence of fibrosis and (3) evaluate correlations between immune-inflammatory markers and fibrosis severity.

A total of 180 patients were categorized as MASLD (n = 69), MetALD (n = 43), and ALD (n = 68) based on the Delphi consensus. SLD was defined as hepatic steatosis on imaging in the presence of either metabolic dysfunction and/or alcohol exposure. Advanced fibrosis was defined as a liver stiffness measurement ≥12 kPa on transient elastography. FIB-4, APRI, Agile 3+ MELD, CTP scores, and systemic immune-inflammatory markers-neutrophil-to-lymphocyte ratio (NLR) and others, were assessed. Correlation analysis and multivariate logistic regression were performed. We derived a composite score using inflammatory markers, which identified patients with increased likelihood of advanced fibrosis, with high accuracy and a high positive predictive value.

Out of 180 patients with SLD, ALD patients had the highest neutrophilic burden (mean NLR 6.56 ± 1.61), followed by MetALD (5.16 ± 1.36) and MASLD (4.30 ± 1.60) (P < 0.001). MASLD exhibited a lymphocyte-predominant profile (mean LMR 4.86 ± 1.17), as compared to MetALD (3.46 ± 0.74) and ALD (3.31 ± 0.98) (P < 0.001). A total of149 patients had advanced fibrosis. Fibrosis burden was highest in ALD (mean liver stiffness 27.43 ± 11.50 kPA), but it was similar in both MetALD (20.28 ± 14.29 kPA) and MASLD (20.49 ± 7.66 kPA).

Despite overlapping clinical features, MASLD, MetALD, and ALD exhibit distinct inflammatory profiles. Systemic immune-inflammatory markers, especially NLR, are closely associated with fibrosis severity. The developed Fibrosis-Inflammation Risk Model Score, offers a potentially scalable, cost-effective tool pending validation in broader real-world settings.