Advances in MR angiography support a growing role for the imaging modality in noninvasive evaluation of the carotid arteries, but surgeons still lack confidence in it, according to data presented at an RSNA refresher course.
Three-dimensional MRA provides for faster imaging with minimal loss in image clarity. Occurrence of artifacts has also been reduced, in comparison to 2-D imaging. Total exam time has decreased 40% to 45%, primarily due to faster reconstruction, said Dr. Andrew Litt.
"For MRA without contrast, the 3-D technique is now favored at my institution," Litt said. "We find that it offers a number of advantages over 2-D imaging."
In particular, 3-D provides better images for distinguishing occlusive versus nonocclusive lesions at the carotid bifurcation. The distinction is critical because surgeons will not operate, at least not immediately, on most total occlusions, as tissue damage has already occurred. On the other hand, a severe but nonocclusive lesion poses a near-urgent surgical situation.
"In the case of a nonocclusive lesion, surgery can have a major impact on outcome by restoring flow," Litt said.
Use of contrast can further improve MRA evaluation of the bifurcation, said Dr. Patrick Turski, a radiologist at the University of Wisconsin. Contrast overcomes problems related to saturation of slow flow and associated loss of signal intensity, as well as difficulties posed by turbulent flow.
Contrast results in venous enhancement, which can at least partially obscure the bifurcation, Turski said. But the problem can be offset by sampling low spatial frequencies more often than high spatial frequencies.
Radiologists can rely on "a number of tricks," especially maneuvers involving digital subtraction techniques, to correct image problems with MRA, he added.
Despite the resulting improvement in image quality with MRA, surgeons still prefer conventional angiography for imaging of the carotid circulation, according to Litt. A study conducted by the American Society of Neuroradiologists provided some disturbing data about surgeon's perceptions of MRA, he said.
The ASNR study compared MRA and conventional angiography in several dozen patients who had symptoms consistent with carotid stenosis. Surgeons were given the results of MRA studies, followed by carotid angiography findings.
"The primary objective of the study was to determine the effect of the imaging modality on decision-making," Litt said. "Does the image provide enough data so that the surgeon feels comfortable making a decision about proceeding with surgery?"
From a technical standpoint, the study showed good correlation between MRA and angiography. For stenoses greater than 60%, MRA had 98% sensitivity and 72% specificity. The figures were 90% and 80% for sensitivity and specificity, respectively, for stenoses of 70% or greater.
Surgeons expressed significantly greater confidence in conventional angiography, however. On a scale of one to five, MRA was rated about 3.5, as compared to 4.5 for angiography. Of even greater concern, surgeons changed their treatment decisions in seven patients after review angiographic results, reversing decisions they had initially made on the basis of MRA.
"That was a particularly disturbing finding, and it made us realize that the radiology community needs to step back and look at what we're doing with MRA and find out what needs to be done to address surgeons' concerns," Litt said.
