Transient Hepatic Elastography and Detection of Clinically Significant Esophageal Varices

Abstract

Introduction: Esophageal varices (EV) are a complication
of portal hypertension, and its rupture is associated with
a high mortality rate. The use of non-invasive diagnostic
methods, including hepatic elastography, have been under
investigation.
Aim: Evaluate the usefulness of transient hepatic elastography
as a non-invasive test for the prediction of presence
of EV, to avoid systematic screening with esophagogastroduodenoscopy
in a selected group of patients.
Material and Methods: Retrospective analysis of the patients
of our Hepatology Unit with a hepatic elastography value >13
kPa who underwent esophagogastroduodenoscopy within
18 months of its evaluation.
Results: We included 259 patients (80.7% males), with a
medium age of 57.9 ± 9.8 years; 65.3% had EV (52.1%
grade II-III). We built ROC curves correlating hepatic elastography
value and presence of EV, and obtained an area
under the curve (AUC) for presence of EV of 0.755 (cut-off
>26.6 kPa, sensitivity 81.1%, specificity 64.4%) and AUC
of 0.68 (cut-off > 28,1 kPa, sensitivity 82.9%, specificity
49.7%) for grade II-III EV. For viral cirrhosis we obtained
AUC of 0.83 for presence of EV (cut-off > 19.8 kPa, sensitivity
95.4%, specificity 62.2%) and AUC of 0.69 for grade
II-III EV (cut-off > 19.8 kPa, sensitivity 94.7%, specificity
42.0%). In patients without splenomegaly or thrombocytopenia,
AUC was 0.89 (cut-off > 26.6 kPa, sensitivity 86.3%,
specificity 81.8%) for presence of EV and 0.75 (cut-off >
26. 6kPa, sensitivity 90.9%, specificity 60.6%) for grade IIIII EV.
Conclusion: Hepatic elastography values and presence of
EV were related, with a higher significance in the groups of
viral cirrhosis and those without splenomegaly or thrombocytopenia. Cut-off points had high sensitivity, illustrating the
potential utility of transient hepatic elastography in identifying
patients who may not need to undergo esophagogastroduodenoscopy.