US-FLI score - Is it possible to predict the steatosis grade with an ultrasonographic score?

Abstract

Objectives: A recent ultrasonographic score (Ultrasonographic fatty liver indicator (US-FLI)) allows to grade steatosis severity on ultrasound (US).We aimed to evaluate the agreement of US-FLI with the controlled attenuation parameter (CAP) in patients with non-alcoholic fatty liver disease (NAFLD).

Methods: Initially, inter-observer agreement for the score was assessed between 3 physicians using a sample of 31 patients.Later, 96 patients with NAFLD were included and several anthropometric/clinical/analytical parameters were assessed and US and transient elastography was performed.

Results: Physicians showed an excellent absolute agreement regarding the total score, with an average Interclass Correlation Coefficient of 0.972(95% CI 0.949-0.986). Comparing US-FLI with CAP, considering the previously defined cut-off for steatosis >S1(268dB/m) and > S2(280dB/m), US-FLI had a good discriminative capacity for both grades, with areas under the curve (AUC) of 0.88(p < 0.001) and 0.90(p < 0.001), respectively.Also, US-FLI ≤ 3 points had a negative predictive value of 100% for steatosis >S2 and US-FLI ≥6 points had a positive predictive value (PPV) of 94.0% for steatosis >S2. When comparing the clinical score Fatty Liver Index (FLI) for the same CAP cut-offs, it showed a weak discriminative capacity for both grades, with AUC of 0.65(p = 0.030) and 0.66(p = 0.017). AUC for US-FLI and FLI were significantly different for both cut-offs (p < 0.001).

Conclusion: US-FLI has an excellent reproducibility and a good discriminative capacity for the different steatosis grades.Scores ≤3points exclude significant steatosis and scores ≥6 points have a PPV of 94,0% for steatosis >S2.US-FLI was significantly superior to the clinical score FLI in the discrimination between steatosis grades.

Keywords: Controlled attenuation parameter; Non-alcoholic fatty liver disease; Steatosis; Ultrasonography.