MR elastography in patients with suspected diffuse liver disease at 1.5T: Intraindividual comparison of gradient-recalled echo versus spin-echo echo-planar imaging sequences and investigation of potential confounding factors. Issue 142 (September 2021)

• Record Type: Journal Article
• Authors: Plaikner, Michaela; Kremser, Christian; Zoller, Heinz; Kannengiesser, Stephan; Henninger, Benjamin
• Is Part Of: European journal of radiology. Issue 142(2021)
• Publication Date: 2021-09
• Subjects: MR elastography -- Iron overload -- Steatosis -- Liver; MRE Magnetic resonance elastography -- GRE-MRE Gradient-recalled echo MRE -- SE-EPI-MRE Spin-echo echo-planar imaging MRE -- hiSE-EPI-MRE High-iron spin-echo echo-planar imaging MRE -- LS Liver stiffness -- ROI Region of interest -- CCC Concordance correlation coefficient -- SVS Single-voxel STEAM spectroscopy -- PDFF Proton density fat fraction -- IQR Interquartile range -- SEM standard error of the mean -- LOA limits of agreement; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109898
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: For diffuse liver disease, GRE- and SE-EPI-MRE stiffness values show good agreement at 1.5T. The number of successful exams for different MRE sequences is mainly determined by iron content. SE-EPI-MRE enables usable MRE results in patients with iron overload for R2* up to 563 1/s. For patients with iron overload SE-EPI-MRE should be preferred. Abstract: Purpose: To compare liver stiffness (LS) in patients with suspected diffuse liver disease between gradient-recalled-echo magnetic resonance elastography (GRE-MRE) and different spin-echo echo-planar imaging (SE-EPI-MRE) sequences and to investigate confounding factors including fat, iron, age, and sex. Method: LS was measured at 1.5T using GRE-MRE, SE-EPI-MRE and short-TE-SE-EPI-MRE (hiSE-EPI-MRE) sequences and compared using Bland-Altman-plots together with concordance correlation coefficients (CCC). Success gradings were evaluated considering possible confounding factors. Results: 305 patients (225 male, 80 females, mean age 51.12 years) were included. 109/305 showed hepatic iron overload, 183 hepatic steatosis. The mean difference (bias) in stiffness values between GRE-MRE and SE-EPI-MRE/hiSE-EPI-MRE was 0.15/0.2 kPa (LOA: −0.72, 0.41 kPa/-0.94, 0.55 kPa), between SE-EPI-MRE and hiSE-EPI-MRE 0.04 kPa (LOA: −0.62, 0.53 kPa). The CCC for agreement between stiffness values for GRE-MRE and SE-EPI-MRE was 0.94 (0.92–0.95), 0.89 (0.86–0.91) for hiSE-EPI-MRE and GRE-MRE and 0.94 (0.92–0.95) for SE-EPI-MRE and hiSE-EPI-MRE. Using GRE-MRE, 72/305 showed unusable results whereby all these patients had high iron levels (mean R2*=209.7 1/s). For SE-EPI-MRE and hiSE-EPI-MRE only 10/305 and 8/305 were inconclusive respectively, corresponding to a significantly higher iron load (mean R2*= 549.2 1/s for SE-EPI-MRE and 570.7 1/s for hiSE-EPI-MRE). Concerning fat, age or sex no significant influence on success was observed for all sequences. Conclusions: Good agreement of LS values was observed between GRE-MRE and SE-EPI-MRE sequences. The number of successful exams, however, was considerably lower for GRE-MRE, mainly due to iron content. Study reference number: AN5093.