Usefulness of controlled attenuation parameter in monitoring clinically relevant decline of hepatic steatosis for non-alcoholic fatty liver disease.

Abstract

Magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) is the reference standard of hepatic steatosis assessment. This study evaluates usefulness of controlled attenuation parameter (CAP) in monitoring the clinically relevant outcome by MRI-PDFF for non-alcoholic fatty liver disease (NAFLD) patients. NAFLD patients were enrolled prospectively. Instruction was given in lifestyle modifications with exercise and control of metabolic factors. MRI-PDFF and CAP were performed at enrollment and follow-up, with the diagnostic validity of CAP in monitoring clinically relevant outcome defined as a decline of ≥30% relative to baseline value by MRI-PDFF. A total of 75 patients (male/female: 49/26, mean age: 53.2) were enrolled. Baseline MRI-PDFF, CAP and liver stiffness was 14.4%, 300.2 dB/m and 6.5 kPa. In a median interval of 369 days, thirteen (17.3%) patients achieved clinically relevant outcome with decline of 46.7 dB/m by CAP, compared with increase of 5.1 in the other patients. In multivariate analysis, clinically relevant outcome was associated with changes (Δ) of CAP and glucose. Assessed by area under receiver operating curve, the performances of ΔCAP in predicting clinically relevant outcome were 0.815 and 0.808, and with the specificity of >90%, the ΔCAP cutoff was -46 dB/m and -15% relative to baseline value; sensitivity was 53.8% and 46.2% with negative predictive value of 90.3% and 88.9% respectively. For NAFLD patients, CAP exhibited good performance in monitoring clinically relevant decline of hepatic steatosis in MRI-PDFF. With the cutoffs of -46 dB/m or -15%, ΔCAP is useful in excluding clinical relevant outcome achievement.