Breath-hold MR cholangiography with snapshot techniques: prospective comparison with endoscopic retrograde cholangiography.

Abstract

To compare findings with magnetic resonance (MR) cholangiography with rapid acquisition with relaxation enhancement (RARE) and half-Fourier acquisition with single-shot turbo spin-echo (hereafter, half Fourier RARE) snapshot imaging techniques to those with endoscopic retrograde cholangiography (ERC). Heavily T2-weighted thick-section (RARE) and thin-section (half-Fourier RARE) MR cholangiography were performed prospectively, on a 1.5-T imager, in the biliary tree of 61 consecutive patients before ERC. Findings at ERC were considered the standard of reference. The radiologist and endoscopist were blinded to each other's report. On- and off-site MR cholangiographic readings were performed to detect stones (n = 24), biliary dilatation (n = 34), or stenosis (n = 36). The sensitivity and specificity of MR cholangiography, respectively, calculated on a lesion-by-lesion basis, were 92.3% and 95.8% for cholangiolithiasis, 94.1% and 92.6% for duct dilatation, and 88.8% and 84.0% for stenosis. With snapshot MR cholangiography, on a patient-by-patient basis, differentiation between normal (n = 15) and abnormal (n = 46) results yielded a sensitivity of 92.4%, a specificity of 83.4%, and a positive predictive value of 95.6%. Pitfalls were caused by flow artifacts, compression by vessels, and low contrast between calculi and surrounding parenchyma. Snapshot MR cholangiography allowed noninvasive, accurate detection of biliary stones, strictures, and dilatation similar to that with ERC. Discrepancies regarding low-grade dilatation and strictures had no clinical relevance at retrospective review.