Correlation of the grade of hepatic steatosis between controlled attenuation parameter and ultrasound in patients with fatty liver: a multi-center retrospective cohort study

Background/Aims The controlled attenuation parameter (CAP), based on transient elastography, is widely used for noninvasive assessment of the degree of hepatic steatosis (HS). We investigated the correlation of the degree HS between CAP and ultrasound (US) in patients with HS. Methods In total, 986 patients with US-based HS who underwent transient elastography within 1 month were evaluated. The US-based grade of HS was categorized as mild (grade 1), moderate (grade 2), or severe (grade 3). Results The CAP was significantly correlated with the US-based grade of HS (r = 0.458, p < 0.001). The median CAP value of each US-based HS grade showed a positive correlation with grade (271.1, 303.7, and 326.7 dB/m for grades 1, 2, and 3). In a multivariate analysis, the US-based HS grade, body mass index, serum albumin, alanine aminotransferase, and total cholesterol, and liver stiffness were all significantly correlated with the CAP value (all p < 0.05). The areas under the receiver operating characteristic curves for grade 2 to 3 and grade 3 HS were 0.749 (95% confidence interval [CI], 0.714 to 0.784) and 0.738 (95% CI, 0.704 to 0.772). The optimal cut-off CAP values to maximize the sum of the sensitivity and specificity for grade 2 to 3 and grade 3 HS were 284.5 dB/m (sensitivity 78.6%, specificity 61.7%) and 298.5 dB/m (sensitivity 84.6%, specificity 55.6%). Conclusions The correlation of the degree of HS between CAP and US was significantly high in patients with HS, and the optimal cut-off CAP values for grade 2 to 3 and grade 3 HS were 284.5 and 298.5 dB/m.