Utilizing CT-derived extracellular volume fraction to distinguish hepatocellular adenoma from focal nodular hyperplasia

Abstract

This study evaluates the efficacy of computed tomography derived extracellular volume (ECV) fraction in differentiating focal nodular hyperplasia (FNH) from hepatocellular adenoma (HCA). Forty-eight pathologically proven lesions (21 HCA and 27 FNH) in 41 patients (18 with HCA and 23 with FNH) were retrospectively examined in this single-center study conducted between March 2015 and September 2023. The ECV fraction was calculated from unenhanced and equilibrium phase computed tomography images. Receiver operating characteristic curve analysis was used to determine the best discrimination between FNH and HCA. The study population consisted of patients diagnosed with HCA (16.7% male, mean age 34.2 +/- 10.3 years) and FNH (26.1% male, mean age 41.6 +/- 9 years). The ECV fraction of the FNH group was significantly higher than that of the HCA group (37.7%+/- 8.8% vs 26.7% [interquartile range: 22.5%-31%], P = .001). The optimal cutoff value for the ECV fraction in differentiating FNH from HCA was determined to be 32.25%, with 76.5% sensitivity and 78.9% specificity (area under the curve, 0.824; 95% confidence interval [CI]: 0.675-0.972, P = .001). The interobserver reliability of ECV fraction measurement was excellent for both HCA (intraclass correlation coefficient, 0.884; 95% CI: 0.724-0.953, P = .001) and FNH (intraclass correlation coefficient, 0.915; 95% CI: 0.764-0.969, P < .001). ECV fraction demonstrated high performance and efficacy as a potential diagnostic tool for differentiating FNH from HCA.