Assessment of acute cholangitis by MR imaging.

PURPOSE: The purpose of this study is to assess the common MRI findings of acute cholangitis compared with those of non-acute cholangitis.

MATERIALS AND METHODS: During a 31-month period, we performed MRCP and contrast-enhanced MRI on 173 patients with biliary abnormalities including duct dilatation or stricture. The causes of the biliary abnormalities included biliary stone disease (n=85), cholangiocarcinoma (n=47), periampullary cancer (n=20), GB cancer (n=4), and others (n=17). Among 173 patients, 66 consecutive patients were confirmed with acute cholangitis diagnosed according to the Tokyo guideline, and 107 patients were confirmed as having non-acute cholangitis. Two radiologists retrospectively and independently accessed the MR findings, including the cause of biliary abnormality, increased periductal signal intensity on T2-weighted images, the transient periductal signal difference, and the presence of abscess, thrombosis, and ragged duct. They also measured the dilated duct and the thickened wall. The Student t-test and the Pearson chi-square were used. The κ statistics were used to determine interobserver agreement. Logistic regression was used to identify the MR findings that predicted acute cholangitis.

RESULTS: MRI correctly accessed the cause of biliary abnormality in 163 patients (94%). The statistically common findings for acute cholangitis were as follows: increased periductal signal intensity on T2-weighted imaging (n=26, 39%, p<0.05); transient periductal signal difference (n=31, 47%, p<0.05); abscess (n=18, 27%, p<0.05); thrombosis (n=12, 18%, p<0.05); and ragged duct (n=11, 17%, p<0.05). Interobserver agreement was good to excellent for each finding (κ=0.74-0.97). The wall thickness showed a statistically significant difference between the acute cholangitis and the non-acute cholangitis group (2.65 mm:2.32 mm, p<0.05), however, there was no significant difference in duct dilatation in the two groups. The periductal transient attenuation difference was an independent predictor of acute cholangitis (Exp (B)=6.389, p=0.018).

CONCLUSION: MRI accurately assesses the cause of biliary abnormality in patients with cholangitis. Using statistically common MR findings for acute cholangitis, MR imaging is very successful in predicting acute cholangitis.