BACKGROUND/ PURPOSE: This study was undertaken to assess the usefulness of magnetic resonance angiography (MRA) in detecting varices and to investigate the possibility of avoiding routine endoscopy after surgery for biliary atresia (BA). METHODS: The subjects are 21 patients who have undergone periodical MRA since 1996. The esophageal and gastric vessels were investigated as the sites of potential development of collateral vessels. The collateral vessels on MRA were compared with endoscopic findings. RESULTS: (1) Detection of collateral vessels: of 21 patients examined, 9 patients had 12 varices. MRA depicted collateral vessels in all of the 12 varices (sensitivity, 100%). However, all MRA findings, but one was compatible with endoscopic findings (specificity, 92.9%). Outside of a series of these 21 patients, there were 3 additional patients who had undergone endoscopic treatments previously with success, and in whom esophageal vessels were not depicted on MRA, but endoscopy showed remnant varices. (2) Time-dependent relationship between the appearance of collateral vessels and varices: of 4 varices that appeared after the start of MRA, 3 esophageal varices were found endoscopically simultaneously with or after delineation of collateral vessels on MRA. In the remaining patient, varices were found enodscopically 6 months before the MRA delineation, because a simultaneous MRA was not performed in this case. All these varices were in the early stage. CONCLUSIONS: MRA was highly sensitive and specific in detecting esophagogastric collateral vessels of the portal venous system. An endoscopic examination is unnecessary until collateral vessels are seen on MRA after surgery for BA. Copyright 2001 by W.B. Saunders Company.
BACKGROUND/ PURPOSE: This study was undertaken to assess the usefulness of magnetic resonance angiography (MRA) in detecting varices and to investigate the possibility of avoiding routine endoscopy after surgery for biliary atresia (BA). METHODS: The subjects are 21 patients who have undergone periodical MRA since 1996. The esophageal and gastric vessels were investigated as the sites of potential development of collateral vessels. The collateral vessels on MRA were compared with endoscopic findings. RESULTS: (1) Detection of collateral vessels: of 21 patients examined, 9 patients had 12 varices. MRA depicted collateral vessels in all of the 12 varices (sensitivity, 100%). However, all MRA findings, but one was compatible with endoscopic findings (specificity, 92.9%). Outside of a series of these 21 patients, there were 3 additional patients who had undergone endoscopic treatments previously with success, and in whom esophageal vessels were not depicted on MRA, but endoscopy showed remnant varices. (2) Time-dependent relationship between the appearance of collateral vessels and varices: of 4 varices that appeared after the start of MRA, 3 esophageal varices were found endoscopically simultaneously with or after delineation of collateral vessels on MRA. In the remaining patient, varices were found enodscopically 6 months before the MRA delineation, because a simultaneous MRA was not performed in this case. All these varices were in the early stage. CONCLUSIONS: MRA was highly sensitive and specific in detecting esophagogastric collateral vessels of the portal venous system. An endoscopic examination is unnecessary until collateral vessels are seen on MRA after surgery for BA. Copyright 2001 by W.B. Saunders Company.