Transient elastography and ultrasonography for the prediction of liver fibrosis in infants with biliary atresia

Abstract

Purpose: To compare transient elastography (TE) and sonographic findings, such as triangular cord (TC) thickness and hepatic artery (HA) and portal vein (PV) diameters, with histologic fibrosis stage for assessing liver fibrosis in infants with biliary atresia (BA).Materials and Methods: Between April 2007 and July 2010, 51 infants with BA who underwent both TE and ultrasonography (US) before surgery or liver biopsy were retrospectively included. Four patients who had less than 5 valid shots on TE were excluded. Thus 47 of 51 patients (19 boys, median age 60 days) were analyzed. The TC thickness and diameters of the HA and PV were evaluated on US. TE measurements and US findings were compared with the METAVIR fibrosis stage [F0 indicated no fibrosis (n=0); F1, portal fibrosis without septa (n=1); F2, portal fibrosis with a few septa (n=27); F3, numerous septa without cirrhosis (n=14); and F4 cirrhosis (n=5)]. The diagnostic value of TE and US were evaluated. We used the M probe, a probe for general adult population, until June 2009 and the S probe, a specific probe for children, after July 2009. The effect of the M probe and the S probe on diagnostic accuracy of TE was also assessed.Results: Only TE (r=0.63; p<0.001) was significantly correlated with the METAVIR fibrosis stage in infants with BA. Areas under the receiver operating characteristic curve (Az) of TE were 0.86 and 0.96 for the diagnosis of severe fibrosis (≥F3) and cirrhosis (F4), respectively. Cut-off values of TE measurements were >9.6 kPa (sensitivity 89.5%/specificity 75%) and >18.1 kPa (sensitivity 100%/specificity 90.5%) for the diagnosis of severe fibrosis (≥F3) and cirrhosis (F4), respectively. Success rate of the S probe (100%) was significantly higher than that of the M probe (77%, p<0.001). Diagnostic performance of the S probe (0.93) tends to be increased compared to the M probe (0.85) in predicting severe fibrosis (≥F3), but not significant. The S and the M probe showed compatible diagnostic accuracy in predicting cirrhosis (F4; 0.96 and 0.94, respectively).Conclusion: TE may be a useful, noninvasive method for the diagnosis of severe fibrosis (≥F3) and cirrhosis (F4) in infants with BA. Success rate of TE may be improved when using the S probe, a specific probe adapted for children. Further study is needed for the evaluation of the effect of the S probe on diagnostic accuracy of TE in assessing the degree of liver fibrosis in infants with BA.