RESUMO
PURPOSE: Noninvasive evaluation of portal hypertension is needed for cirrhotic patients with large esophageal varices. This study was aimed at assessing the diagnostic value of liver/spleen stiffness in predicting hepatic vein pressure gradient (HVPG) in this special population. METHODS: In the present prospective cohort study, liver/spleen stiffness was measured by transient elastography. Patients also underwent HVPG assessment, upper gastrointestinal endoscopy, and other noninvasive serum models. RESULTS: Ninety-nine cirrhotic patients with large esophageal varices were enrolled. Liver/spleen stiffness strongly correlated with HVPG. In regards to significant portal hypertension, area under receiver operating characteristic curves (AUROCs) for liver/spleen stiffness were 0.74 and 0.91. Accuracy for detecting significant portal hypertension was 79% for spleen stiffness of 48.9 kPa (sensitivity: 76%, specificity: 100%, positive predictive value: 100%, negative predictive value: 38%) and 75% for liver stiffness of 16.0 kPa (sensitivity: 78%, specificity: 54%, positive predictive value: 92%, negative predictive value: 27%). Similarly, spleen stiffness had significant higher AUROCs for predicting HVPG ≥16 and ≥20 mm Hg than that of liver stiffness and other noninvasive serum models. CONCLUSION: In cirrhotic patients with large esophageal varices, liver stiffness and spleen stiffness correlate with HVPG, and spleen stiffness is superior to liver stiffness in predicting portal hypertension.