OBJECTIVE: The aim of this study was to investigate the impact of portal hypertension (PH) on gadoxetic acid-enhanced liver magnetic resonance imaging (MRI) and assess diagnostic and prognostic implications in comparison to established imaging features of PH. MATERIALS AND METHODS: Institutional review board-approved retrospective study of 178 patients (142 men; median age, 59.4 years) with chronic liver disease undergoing MRI and hepatic venous pressure gradient (HVPG) measurement between January 2008 and April 2015. Magnetic resonance imaging was assessed for established features of PH (splenic and portal vein diameters, portosystemic collaterals, ascites) and for features on 20 minutes delayed T1-weighted gadoxetic acid-enhanced MRI, that is, relative liver enhancement (RLE), biliary contrast excretion, or portal vein hyperintensity or isointensity (ie, portal vein hyperintensity sign, PVHS). Statistics encompassed linear regression, logistic regression, and survival analysis. RESULTS: There was an inverse correlation between HVPG and RLE (r = 0.18, P < 0.0001). On univariate analysis, clinically significant PH (ie, HVPG ≥ 10 mm Hg, n = 109) and severe PH (ie, HVPG ≥ 12 mm Hg, n = 99) were associated with delayed biliary contrast excretion (n = 33) and the PVHS (n = 74) (P < 0.01 for all). Multivariate analysis demonstrated significant associations between the PVHS and severe PH (odds ratio [OR], 3.33; P = 0.008), independently of spleen size (OR, 1.26; P = 0.002), portosystemic collaterals (n = 81; OR, 5.46; P = 0.0001), and ascites (n = 88; OR, 3.24; P = 0.006). Lower RLE and the PVHS were associated with lower 3-year, transplantation-free survival (hazards ratios, 0.98 and 3.99, respectively, P = 0.002 for all), independently of the Child-Pugh and Model for End-Stage Liver Disease scores. CONCLUSIONS: The presence of the PVHS on gadoxetic acid-enhanced MRI is an independent indicator of severe PH and may enable more accurate diagnosis. This feature and decreased hepatic contrast uptake may also comprise prognostic information.
OBJECTIVE: The aim of this study was to investigate the impact of portal hypertension (PH) on gadoxetic acid-enhanced liver magnetic resonance imaging (MRI) and assess diagnostic and prognostic implications in comparison to established imaging features of PH. MATERIALS AND METHODS: Institutional review board-approved retrospective study of 178 patients (142 men; median age, 59.4 years) with chronic liver disease undergoing MRI and hepatic venous pressure gradient (HVPG) measurement between January 2008 and April 2015. Magnetic resonance imaging was assessed for established features of PH (splenic and portal vein diameters, portosystemic collaterals, ascites) and for features on 20 minutes delayed T1-weighted gadoxetic acid-enhanced MRI, that is, relative liver enhancement (RLE), biliary contrast excretion, or portal vein hyperintensity or isointensity (ie, portal vein hyperintensity sign, PVHS). Statistics encompassed linear regression, logistic regression, and survival analysis. RESULTS: There was an inverse correlation between HVPG and RLE (r = 0.18, P < 0.0001). On univariate analysis, clinically significant PH (ie, HVPG ≥ 10 mm Hg, n = 109) and severe PH (ie, HVPG ≥ 12 mm Hg, n = 99) were associated with delayed biliary contrast excretion (n = 33) and the PVHS (n = 74) (P < 0.01 for all). Multivariate analysis demonstrated significant associations between the PVHS and severe PH (odds ratio [OR], 3.33; P = 0.008), independently of spleen size (OR, 1.26; P = 0.002), portosystemic collaterals (n = 81; OR, 5.46; P = 0.0001), and ascites (n = 88; OR, 3.24; P = 0.006). Lower RLE and the PVHS were associated with lower 3-year, transplantation-free survival (hazards ratios, 0.98 and 3.99, respectively, P = 0.002 for all), independently of the Child-Pugh and Model for End-Stage Liver Disease scores. CONCLUSIONS: The presence of the PVHS on gadoxetic acid-enhanced MRI is an independent indicator of severe PH and may enable more accurate diagnosis. This feature and decreased hepatic contrast uptake may also comprise prognostic information.
Authors: Ute Lina Fahlenkamp; Katharina Ziegeler; Lisa Christine Adams; Sarah Maria Böker; Günther Engel; Marcus Richard Makowski Journal: Sci Rep Date: 2020-10-22 Impact factor: 4.379