[Usefulness of liver stiffness measurement for predicting the presence of esophageal varices in patients with liver cirrhosis]

Korean J Hepatol. 2008 Sep;14(3):342-50. doi: 10.3350/kjhep.2008.14.3.342.
[Article in Korean]

Abstract

Background/aims: Bleeding from esophageal varices (EV) is a major cause of death in patients with liver cirrhosis. Endoscopic screening is recommended for diagnosing EV, but various noninvasive parameters can also be used to predict EV. The liver stiffness measurement (LSM), a noninvasive technique for estimating liver fibrosis, was recently reported to be strongly correlated with the hepatic venous pressure gradient. This study evaluated the usefulness of LSM for predicting the presence and size of EV in patients with cirrhosis.

Methods: The relationships of LSM with the presence and size of EV were analyzed in 112 patients with liver cirrhosis. Liver cirrhosis was diagnosed histologically or clinically. The presence and size of EV were assessed by endoscopy, and LSM was determined by the Fibroscan technique.

Results: LSM was strongly correlated with the presence of EV (P<0.0001): the LSM value was 42.7+/-21.9 kPa (mean+/-standard deviation) in patients with EV (n=82) and 19.1+/-12.6 kPa in patients without EV (n=30). The area under the receiver operating characteristic curve was 0.818 (95% CI, 0.732-0.904) for predicting the presence of EV, and an LSM value of 19.7 kPa was predictive of the presence of EV with a sensitivity of 87%, a specificity of 70%, a PPV of 89%, and a NPV of 66%. However, there was a weak correlation between LSM and the size of EV.

Conclusions: LSM is useful for predicting the presence of EV in patients with cirrhosis but not their size.

MeSH terms

  • Adult
  • Aged
  • Elasticity
  • Esophageal and Gastric Varices / diagnostic imaging*
  • Esophageal and Gastric Varices / etiology
  • Female
  • Hepatic Veins
  • Humans
  • Liver / diagnostic imaging*
  • Liver Cirrhosis / complications*
  • Liver Cirrhosis / physiopathology
  • Male
  • Middle Aged
  • Portal Pressure
  • Predictive Value of Tests
  • ROC Curve
  • Severity of Illness Index
  • Ultrasonography