Correlation and discriminatory performance of non-invasive serum fibrosis indices according to liver disease etiology...

Non-invasive serum fibrosis indices are widely used for liver fibrosis assessment; however, their correlation with liver stiffness and diagnostic performance may vary according to liver disease etiology. This study aimed to evaluate the etiology-dependent correlation between multiple serum Non-Invasive Test (NITs) indices and liver stiffness measured by two-dimensional Shear Wave Elastography (2D-SWE), and to explore their ability to discriminate patients according to 2D-SWE-based fibrosis categories.

We conducted a retrospective cross-sectional study including 105 patients with chronic liver disease. Serum NITs indices were correlated with liver stiffness measured by 2D-SWE using Pearson or Spearman correlation Exploratory ROC curve analyses were performed to assess the ability of serum indices to discriminate 2D-SWE-based fibrosis categories by etiology.

Strong intercorrelations were observed among multiple serum NITs indices across all etiological groups, indicating substantial collinearity. In metabolic-associated steatotic liver disease, alcohol-related liver disease, and metabolic dysfunction-associated alcohol-related liver disease, correlations between serum indices and liver stiffness were weak to moderate. In contrast, viral hepatitis showed moderate and statistically significant correlations between several indices and 2D-SWE, with more consistent diagnostic performance in ROC analyses. Autoimmune hepatitis demonstrated an intermediate pattern, with modest but significant associations between selected indices and liver stiffness. Exploratory ROC analyses showed heterogeneous discriminatory performance across etiologies.

Serum NITs indices exhibit high collinearity and etiology-dependent performance. Their correlation with liver stiffness and their discriminatory performance against 2D-SWE-based fibrosis categories vary according to underlying liver disease supporting the complementary rather than interchangeable use of serum indices and elastography.