Validation trial for efficacy of ultrasonographic measurement method to predict ascitic volume using virtual ultrasonography.

Abstract

The aim of this study was to clarify whether ultrasound quantitative methods were positively correlated with volume of ascites evaluated by whole abdominopelvic CT. Sixty-eight patients with cirrhotic ascites were retrospectively analyzed. First, to confirm that virtual ultrasonography (VUS) is an alternative method to conventional ultrasound, 22 patients underwent both conventional ultrasonography and VUS. Second, the efficacy of US quantitative methods (3-point method, 4-point method, 5-point method, and Matsumoto's method) was confirmed by VUS in 68 patients. We assessed whether the ascites volume predicted by VUS corresponded with that calculated by 3D-CT. Of the 68 patients, 23 patients were analyzed before and after administration of tolvaptan. The predictive volumes calculated by VUS were remarkably relative to those yielded by conventional US. Correlations between exact volume and those measured by VUS were significantly high (3-point method: r = 0.882, p < 0.001; 4-point method: r = 0.797, p < 0.001; 5-point method: r = 0.836, p < 0.001; Matsumoto's method: r = 0.453, p < 0.001). Correlations between decreasing volume on 3D-CT and that measured by VUS were also significantly high in patients with administration of tolvaptan. Ascites volume measured by ultrasound was effective, especially the 3-point and 5-point methods. It was useful to assess the efficacy of diuretics in cirrhotic patients.