| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorial |
1 From the Department of Radiology, Brigham and Women's Hospital, 75 Francis St, Boston, MA 02115. Received October 7, 1999; revision requested and received November 18; accepted November 19. Address reprint requests to the author.
Index terms: Editorials • Coronary angiography, comparative studies, 54.12142, 54.1244 • Coronary vessels, diseases, 54.762 • Coronary vessels, MR, 54.121412, 54.121415, 54.12142 • Coronary vessels, stenosis or obstruction, 54.762 • Magnetic resonance (MR), three-dimensional, 54.12142 • Magnetic resonance (MR), vascular studies, 54.12142
In this issue of Radiology, Sardanelli et al (1) have shown very nicely the clinical applicability of the three-dimensional (3D), navigator-echo magnetic resonance (MR) imaging pulse sequence for the evaluation of coronary arterial disease. This article is not the first one describing the use of MR coronary angiography, nor is it likely to be the last. The incremental value of the data presented by Sardanelli et al (1) can be judged best with a brief review of the recent history of MR coronary angiography.
Before the introduction of the navigator-echo sequence, the state of the art for MR coronary angiography was the acquisition of MR arteriographic sections through the coronary tree during voluntary breath holds (2). The technique performed quite well. The early articles (2,3) gave impressive results for small patient cohorts. A seminal article published in 1993 created the impression that the technique was ready for prime time and that within 1–2 years it could replace diagnostic coronary angiography (3). In truth, when properly executed, the breath-hold technique yields extremely good quality images of the coronary arteries. This image quality is hard to surpass, even when compared with the quality of images generated with the newer approaches to coronary MR imaging. The attention to technical detail and the need for a large amount of patient cooperation unfortunately have hampered the wide diffusion of this technique of coronary MR imaging.
When introduced, the navigator-echo pulse sequence was promised to increase the applicability of MR coronary angiography (4). Patients who were not good candidates for imaging with the breath-hold techniques became possible candidates for imaging with the navigator-echo sequence. Patients were still poor candidates if they showed abnormalities in their rhythm of breathing or if they had a marked cardiac arrhythmia.
While the target population for MR coronary angiography was broadened, the quality and accuracy of the MR arteriographic technique were not necessarily preserved. Since acquired image information has to be added with the aid of the gating information, some degradation in the inherent resolution of images obtained with the technique might be expected. This slight difference in resolution with the navigator-echo sequence might lead to some loss of accuracy when compared with the resolution with the breath-hold MR arteriographic approaches (5).
How do the data presented by Sardanelli et al (1) match up with our expectations? The number of candidates in whom the sequence was implemented easily is indicated clearly: 39 of 50, or 78%. This may be a conservative estimate, since six individuals were excluded because of concurrent coronary interventions. We can only speculate as to the image quality that could have been achieved in these six individuals. We are therefore left with an estimate of the target population for MR coronary angiography: somewhere between 78% and 90%. We will come back to the six patients who were excluded on the basis of having undergone coronary arterial intervention. The exclusion of these patients, while it introduces a bias, likely makes for a conservative estimate of the accuracy of MR coronary angiography.
But first, how do the accuracy data for the navigator-echo sequence compare with those for the breath-hold approach? With the navigator-echo sequence, the reported sensitivity and specificity are, for the proximal and distal segments of the coronary arteries, 82% (95% CI: 73%, 91%) and 89% (95% CI: 85%, 94%) (1). The sensitivity and specificity of the breath-hold technique are 90% and 92% for imaging of the proximal and midportions of the major epicardial coronary branches (3).
Sardanelli et al (1) suggest that the accuracy could be improved with the use of off-axis and curved-plane reconstruction. Authors of recently published data suggest that this may not be the case (6). An alternate explanation is the limitation inherent in respiratory gating (7). The final image quality achieved by adding images with gating information is less effective as the arterial diameter decreases. Smaller tolerances on the gating intervals might improve image quality. The coronary arterial-to-non–coronary arterial contrast ratio also is degraded preferentially in smaller branches of the coronary artery. Is there a solution? The authors suggest that the use of a contrast agent may be a more effective remedy to this limitation. This is a reasonable expectation (8).
The diagnostic performance of the navigator-echo MR arteriographic sequence for the proximal coronary arteries should be given its just credit. All proximal segments of the major coronary arteries were depicted. The accuracy of MR angiography in these proximal segments was approximately 90%. These results of 3D, navigator-echo MR angiography should be examined from the perspective of using MR angiography as a screening examination for detecting stenosis of 50% or greater narrowing of the proximal coronary arteries. This shows that the accuracy is comparable to and, in fact, better than what is achieved with other screening examinations such as rest-exercise thallium 201 single photon emission computed tomography, or SPECT (sensitivity, 88%; specificity, 77%), or stress echocardiography (sensitivity, 76%; specificity, 88%) (9).
The data presented by Sardanelli et al (1) suggest that 3D, navigator-echo MR coronary angiography is well positioned to become a screening examination for significant coronary arterial disease. It can be applied to 78%–90% of patients and, in this group of patients, can be used to detect focal coronary lesions with an accuracy of 90%. This estimate of 90% is of course applicable to cohorts with disease prevalence comparable with that of the current group: Of 39 patients, 34 examined with MR coronary angiography and with conventional coronary angiography had coronary arterial disease, for a disease prevalence of 87%. A fair evaluation of the true diagnostic performance of MR coronary angiography would require a cohort with a lower prevalence of coronary arterial disease.
The use of data from such a cohort would minimize the overestimation of diagnostic accuracy associated with high disease prevalence and would decrease the verification bias that has been observed for other diagnostic examinations (10). The authors have, perhaps inadvertently, taken a step toward remedying this situation by excluding some patients from their comparison group. The six patients who underwent coronary intervention and the five patients who could not tolerate the MR imaging examination were excluded from the final cohort. These 11 patients all had significant coronary arterial disease. The study composition is therefore slightly less biased toward patients with coronary arterial disease.
This is a step in the right direction, but a fair evaluation of 3D, navigator-echo MR angiography as a screening examination must wait for results from a larger cohort with a stronger proportion of patients with symptoms and with normal coronary arteriograms.
The changing nature of MR imaging technology is such, however, that newer breath-hold approaches promise to improve on the diagnostic performance of MR coronary angiography. Single-breath-hold, multisection MR angiography is now possible and is able to offer good-quality images. The 3D, navigator-echo imaging sequence has the advantage of a relatively simple imaging protocol with minimal requirements for the patient.
The limitations for imaging the distal coronary arterial segments remain, although they might be improved on. The use of contrast agents promises to improve the signal-to-noise ratio in these distal coronary branches and might increase the diagnostic accuracy of 3D, navigator-echo MR angiography. Image-processing time will shorten as new software becomes available.
This exciting evolution in MR coronary angiography makes the reader wonder what the future of MR coronary angiography looks like. The reader can believe easily that more improvements are soon to follow and that the technique has not yet reached full maturity. One must stop and think about the direction taken with once-innovative coronary and cardiac imaging approaches. These technologies—cardiac nuclear medicine, conventional coronary angiography, and echocardiography—were once innovative and were part of our research efforts. With their maturity, they have found a home.
Footnotes
Abbreviation 3D = three-dimensional
See also the article by Sardanelli et al (pp 808–814 ) in this issue.
References
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
