Liver Stiffness Measurements by 2D Shear-Wave Elastography: Effect of Steatosis on Fibrosis Evaluation.

Abstract

<b>BACKGROUND.</b> Hepatic steatosis has been found not to affect liver stiffness measurements (LSM) from MR elastography (MRE). However, the effect of steatosis on LSM from 2D shear-wave elastography (SWE) remains controversial. <b>OBJECTIVE.</b> The purpose of this study was to evaluate the effect of hepatic steatosis on the diagnostic performance of LSM from 2D SWE (LSM_{2D SWE}) for evaluation of liver fibrosis with LSM from MRE (LSM_MRE) as the reference standard. <b>METHODS.</b> This retrospective study included 888 patients (442 women, 446 men; median age, 67 years) with chronic liver disease who underwent LSM by both 2D SWE and MRE within a 3-month window. Steatosis was also assessed on ultrasound examinations by ultrasound-guided attenuation parameter (UGAP) and on MRI examinations by proton density fat fraction (PDFF). Fibrosis stages and steatosis grades were classified according to previously established thresholds. The effect of steatosis on LSM_{2D SWE} was evaluated by Kruskal-Wallis tests with post hoc tests and ROC analysis. <b>RESULTS.</b> LSM_{2D SWE} were significantly higher in patients with severe steatosis than those without steatosis by MRI PDFF among patients with F0 fibrosis (5.5 kPa [IQR, 4.7-6.0 kPa] vs 4.7 kPa [IQR, 4.2-5.5 kPa], <i>p</i> = .009) and F1 fibrosis (6.3 kPa [IQR, 6.0-7.2 kPa] vs 5.9 kPa [IQR, 5.0-6.6 kPa], <i>p</i> = .009). LSM_{2D SWE} were significantly higher in patients with severe steatosis than those without steatosis by UGAP among patients with F1 fibrosis (6.6 kPa [IQR, 5.9-7.3 kPa] vs 5.9 kPa [IQR, 5.1-6.5 kPa], <i>p</i> = .008). Otherwise, LSM_{2D SWE} did not vary significantly across steatosis grades at a given fibrosis stage (all <i>p</i> > .05). Sensitivity and specificity for ≥ F1 fibrosis were 63.8% and 91.5% in patients without versus 60.4% and 80.9% in patients with severe steatosis by MRI PDFF and were 62.4% and 91.5% in patients without versus 72.1% and 78.3% in patients with severe steatosis by UGAP. <b>CONCLUSION.</b> Severe hepatic steatosis may result in overestimation of LSM_{2D SWE} in patients without or with mild steatosis, reducing the specificity of liver fibrosis detection. <b>CLINICAL IMPACT.</b> Assessment of UGAP at 2D SWE may help identify patients in whom LSM_{2D SWE} should be assessed with caution. In patients with no or mild steatosis by 2D SWE and severe steatosis by UGAP, MRE helps provide a more reliable measure of liver fibrosis.