Format

Send to

Choose Destination
See comment in PubMed Commons below
Magn Reson Med Sci. 2013;12(2):77-86. Epub 2013 May 10.

MR-based measurements of portal vein flow and liver stiffness for predicting gastroesophageal varices.

Author information

1
Department of Radiology, University of Yamanashi, 1110 Shimokato, Chuo-shi, Yamanashi, Japan.

Abstract

OBJECTIVES:

We evaluated flow parameters measured by phase-contrast magnetic resonance (MR) imaging (PC-MRI) of the portal venous system and liver stiffness measured by MR elastography (MRE) to determine the usefulness of these methods in predicting gastroesophageal varices (GEV) in patients with chronic liver disease (CLD).

METHODS:

In patients with CLD and controls, we performed PC-MRI on the portal (PV) and superior mesenteric veins; calculated mean velocity (V, cm/s), cross-sectional area (S, mm²), and flow volume (Q, mL/min); and determined markers of liver fibrosis (liver stiffness [kPa]) and aspartate aminotransferase (AST) platelet ratio index [APRI]). We visually assessed GEV and development of collateral pathways of the PV on routine contrast-enhanced dynamic MR imaging and compared patient characteristics, flow parameters, liver stiffness markers, and visual analysis among 3GEV groups, those with mild, severe, or no GEV with reference to endoscopic findings.

RESULTS:

Child-Pugh grade, VPV, SPV, liver stiffness, APRI, and visually identified GEV (visible GEV) differed significantly among the 3 groups (P<0.05). We investigated VPV, SPV, liver stiffness, and visible GEV as independent markers to distinguish patients with and without GEV and examined VPV and visible GEV to predict severe GEV. Visible GEV showed low sensitivity (14 to 30%) and high specificity (98%) for predicting GEV in patients with CLD. A subgroup analysis that excluded cases with collateral pathway demonstrated slightly improved diagnostic performance of VPV and liver stiffness.

CONCLUSIONS:

Portal vein flow parameters and liver stiffness can be useful markers for predicting GEV in patients with CLD.

PMID:
23666158
[Indexed for MEDLINE]
Free full text
PubMed Commons home

PubMed Commons

0 comments
How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for J-STAGE, Japan Science and Technology Information Aggregator, Electronic
    Loading ...
    Support Center