Related To:  Chapter 20. Computed Tomography of the Heart;  Chapter 22. Magnetic Resonance Imaging and Computed Tomography of the Vascular System

View in chapter

Technological innovations have advanced noninvasive coronary artery imaging.¹ The integration of retrospective ECG gating and increased CT gantry rotation speeds have improved the temporal resolution and anatomic coverage of multislice CT. Despite these advances and others, a significant limitation remains, namely, the accurate evaluation of calcified coronary arteries. A recent study by Liu et al using 3D free-breathing coronary MRA² describes an interesting approach to this clinically important problem.

Coronary Calcium and CTA

Electron beam computed tomography (EBCT) was the first CT technology to assess coronary artery calcium, and a significant literature correlating coronary artery disease and prognosis with coronary artery calcium burden matured. ^3,4 While helpful in risk stratification, coronary artery calcification is an important technical limitation of CT angiographic quantitation. Artifacts due to partial volume averaging and “blooming artifacts” encountered in the evaluation of the significantly calcified vessel are particularly challenging. In a recent study of 134 symptomatic patients undergoing 64-slice CT within three months of conventional angiography, the sensitivity, number of evaluable segments and positive predictive values were significantly and negatively affected by increased coronary calcium.⁵ In another study using a 32-slice system, 20% of vessels were excluded from analysis; 50% of these exclusions were due to blooming artifacts from extensive calcifications.⁶ In patients with high calcium scores (median Agatson score 434) studied with a 16-detector row system, 35% of vessels were uninterpretable.⁷ Finally, in a study evaluating the role of 64-slice CT in the pre-operative evaluation of severe aortic stenosis, 23% of coronary artery segments studied (142/600) in a cohort of 40 patients were un-interpretable primarily due to coronary artery calcification.⁸ Clearly, the heavily calcified plaque is problematic for accurate CTA assessment of vessel stenosis.^9,10

MR Coronary Angiography and Coronary Calcification

Advances in MR coronary angiography have also shown promise in the non-invasive evaluation of coronary artery disease. Advances in MR coil and gradient technology have made possible faster imaging times, and techniques in parallel imaging have significantly improved coronary artery evaluation. Navigator sequences permit more robust respiratory gating and allow acquisition of three-dimensional volumetric data, and steady-state free precession (SSFP) sequences result in improved performance. Sensitivity, specificity, negative and positive predictive values were recently reported in the 80-85% range.¹¹

In the aforementioned study by Liu et al,² the complementary role of coronary MRA in the evaluation of patients with coronary calcium seen on coronary CTA is described. The authors studied a cohort of 18 patients undergoing evaluation with conventional catheter, retrospective ECG gated 64-slice CT, and MR angiography. Interestingly, inclusion into the study required at least one calcified plaque and a calcium score of >100. Calcified plaque was defined as either diffuse or nodal. A total of 33 calcified plaques were analyzed; 17 patients with nodal and 16 with diffuse calcification

Coronary MRA was performed using either 6 or 8 element phase array coils, ECG-triggered 3D SSFP with respiratory gating, and either volume targeted or whole heart evaluation. CTA and MRA images were interpreted by three radiologists in a blinded fashion, and were compared to conventional angiography. Coronary MRA and CTA images were graded on a 4 point scale, with 4 graded as excellent. Image quality was judged equivalent between CTA and MRA.

In a segmental analysis, 12/33 of the calcified segments demonstrated significant stenoses on conventional angiography. The sensitivity for detection of these stenoses was similar (0.80 vs. 0.75) between MRA and CTA. However, the specificity was significantly greater for MRA than CTA (0.75 vs. 0.48, p=0.02). Moreover, AUC values on ROC analysis were greater with MRA than CTA (0.83 vs. 0.65, p=0.03). Of note, kappa values were higher with MRA than CTA, indicating improved inter-observer agreement in the evaluation of coronary artery stenosis in the setting of coronary calcium. There are limitations to Liu’s study, most notably its small sample size. Nevertheless, these intriguing data illustrate the potential utility of coronary MRA in the evaluation of coronary artery disease.

How Do MRA and CTA Compare?

Although early results of coronary MRA were promising vis-à-vis image quality and detection of stenoses, sensitivity and specificity suffered owing to an incomplete assessment of distal coronary artery segments.¹² Similar performance between 16-slice CT and three-dimensional respiratory navigator gated MRA (sensitivities 75 vs. 82% and specificities 77% vs. 79%, respectively) were recently reported.¹³ In a recent series of 20 patients, subjective image quality with MRA was slightly inferior to CT, and the number of non-assessable segments was greater with MRA than CTA. However, in assessable segments, diagnostic accuracy was comparable with MRI and CT (87% vs. 95%).¹⁴

Bottom Line

The noninvasive imaging field is replete with debate over the “holy grail”, i.e. the “one stop shop for noninvasive cardiac imaging.” A critical evaluation of the current literature indicates that despite advances in CT technology, the presence of significant calcific disease continues to present challenges for the definitive evaluation of coronary artery stenoses. Similarly, while coronary MRA has made equally impressive strides, studies continue to demonstrate important limitations for a definitive and comprehensive assessment of the coronary artery lumen. In addition, patients with pacemakers, claustrophobia and an inability to breathe consistently will still limit access for diagnostic coronary MRA.