CHICAGO - In a series of 68 patients who underwent virtual endoscopy (VE) for evaluation of the common biliary tract, VE was more effective than maximum intensity projection (MIP) MR for identifying small stones, according to a study presented at this week's RSNA show by a team of French researchers.
"VE is effective for stones of 3 mm or more, but MIP did not identify stones of this size," said Dr. Catherine Roy, professor of radiology at Hopitaux University in Strasbourg, France.
Dr. Susan Wall, professor of radiology and vice chairman for academic affairs at the University of California, San Francisco, added that "virtual imaging for tubular structures is clearly the wave of the future because it is noninvasive." Although the risks associated with traditional endoscopy "are minimal, a noninvasive procedure is clearly safer."
Wall said, however, that in the U.S. the "emphasis has been on virtual colonoscopy rather than assessment of the biliary tract."
Roy reported on 68 patients ranging in age from 28 to 82 years with a mean age of 38.5. Forty patients presented with calculi, 13 with cholangiocarcinoma, 10 with pancreatic adenocarcinoma, and five with benign fibrotic strictures. All patients underwent cholangio-MR as part of a conventional MR sequences protocol.
Examinations were performed on a 1-tesla MR unit (Gyroscan NT, Philips Medical Systems, Andover, MA) using a T2-weighted 3-D turbo spin-echo sequence (TR: 1800 ms, TE eff = 500, ETL: 130, slice thickness 0.6 mm, 80 slices, 256 x 256, FOV 240, TA: 4 min 12 s) in coronal plane with respiratory triggering.
The VE images were reconstructed using a volume-rendering technique and interpreted together with overall examination images by two independent radiologists, blinded to clinical data, with consensus by a third.
The VE images were then compared to MIP reconstructions and source images. Correlations with conventional endoscopy and pathological findings served as a standard of reference.
Roy said the VE was "entirely feasible in all patients and the image quality was excellent." But all MR sequences, as well as reconstructed images, missed stones smaller than 2 mm in size.
She said that VE visualized all stones that were 3 mm or larger, and also found 4 cases of stones equal to 3 mm that were not found on source images and T2-weighted sequences.
Among the 20 calculi, ranging in size from 3 to 8 mm, which were well detected on VE, source images, and conventional T2-weighted sequences, nine were invisible on MIP reconstructions.
Tumoral stenosis presented an irregular and asymmetric pattern of the narrowing on endoluminal views. The morphologic assessment (smooth or irregular surface, location and extension in surface of tumoral filling defects) was better seen on VE than on source images. A smooth surface with abrupt margins was present in all cases of stones, and found in only 5 cases of tumoral process.
VE is, however, not effective when evaluating smooth surfaces, and it could not differentiate a smooth calculus from a smooth polypoid tumor. All cases of benign stenosis presented a regular smooth tapering of the common duct.
Roy concluded that VE is more accurate than MIP and is the best modality for detecting tiny stones. Moreover, it "provides excellent film for evaluating the morphology of the common duct. It allows an interesting evaluation of the morphology of a common duct stenosis."
By Peggy Peck
AuntMinnie.com contributing writer
December 6, 2002
Copyright © 2002 AuntMinnie.com
