Purpose: To investigate the frequency of discordance and to identify factors associated with discordance between ultrasonographic and elastographic grades for assessing hepatic fibrosis in Asian patients with chronic hepatitis B.
Methods: Three hundred thirty-four patients with chronic hepatitis B for which both conventional ultrasonography and liver stiffness measurements using elastography were available were included. Patients were graded as 'normal', 'chronic liver disease', or 'liver cirrhosis' by ultrasonography, and as 'no significant fibrosis', 'significant fibrosis', or 'liver cirrhosis' by elastography, and the results of these two modalities were compared. Logistic regression analyses were performed to identify independent factors associated with discordant results.
Results: Of the 334 patients, 153 (45.8 %), 115 (34.4 %), and 66 (19.8 %) patients were 'normal', 'chronic liver disease', and 'liver cirrhosis', respectively, based on the ultrasonographic grades, and 290 (86.8 %), 29 (8.7 %), and 15 (4.5 %) patients were 'no significant fibrosis,' 'significant fibrosis', and 'liver cirrhosis', respectively, based on the elastographic values. Among them, 173 (51.8 %) showed discordance with respect to severity of hepatic fibrosis. In multivariable analysis, discordance was more frequent in patients with ultrasonographic grades of 'chronic liver disease' [adjusted odds ratio (AOR), 1924; P < 0.001] and 'liver cirrhosis' (AOR, 4498; P < 0.001), whereas patients with an elastographic grade of 'liver cirrhosis' showed a negative association with discordance (AOR, 0.002; P = 0.007).
Conclusion: There was a high rate of discordance between hepatic fibrosis grades determined by ultrasonography and elastography. Considering the accuracy of liver stiffness evaluation by elastography, conventional ultrasonography might overestimate hepatic fibrosis in chronic hepatitis B.
Keywords: Elastography; Liver fibrosis; Ultrasonography.