A retrospective assessment of magnetic resonance cholangiopancreatography in children.

OBJECTIVE

To evaluate the utility of magnetic resonance cholangiopancreatography (MRCP) in children and to compare MRCP with direct cholangiopancreatography (CP).

MATERIALS AND METHODS

We performed an unblinded, retrospective chart review of 32 children (ages 0-18 years, 17 male) who underwent MRCP between January 2002 and June 2005. MRCP, endoscopic retrograde cholangiopancreatography (ERCP), percutaneous fluoroscopic or intraoperative studies of the pancreatobiliary tree, and clinical outcomes were evaluated.

RESULTS

Seventeen (52%) children had MRCP alone, 15 (48%) had both MR and direct CP. MRCP results correlated with other evaluative methods in 14/15 (93%) cases. There was 1 false positive (anomalous pancreatic duct union) and 0 false negatives for anatomic abnormalities. Therapeutic intervention was performed in 7 of 28 children initially evaluated by MRCP (2 sphincter of Oddi dysfunction, 2 choledocholithiasis, 2 primary sclerosing cholangitis, 1 congenital hepatic cysts) and 1 of 4 children initially evaluated by ERCP (primary sclerosing cholangitis). All 17 children initially evaluated by MRCP had no change in clinical status to suggest a missed anatomic lesion or therapeutic opportunity.

CONCLUSIONS

In this retrospective study, MRCP was sensitive and specific in identifying anatomic abnormalities of the pancreatobiliary tree in children. MRCP should be considered before direct CP to evaluate anatomic abnormalities of the pancreatobiliary tree.