Accurate Determination of High-Risk Coronary Lesion Type by Multidetector Cardiac Computed Tomography

Invasive coronary arteriography (CA) has been the standard method for imaging coronary lesions since its beginnings in 1958. Multidetector cardiac computerized tomography (MDCT) has undergone rapid developments over the last decade and now provides a noninvasive means to provide high-resolution imaging of the coronary arteries. It has already been established that the presence of coronary calcification by MDCT is an indicator for coronary artery disease and the quantity of calcification is a strong predictor for coronary events. However, there is still marked heterogeneity in the composition of coronary plaques, and calcification is not necessarily a marker of obstructive coronary disease. Therefore, the visualization and characterization of calcified and noncalcified plaques is important in planning revascularization strategies, whether percutaneous or surgical.
At present, numerous studies have evaluated the accuracy of MDCT to determine percent stenosis.^1–3 However, the ability of MDCT to determine coronary lesion type remains unknown. The primary objective of this study was to determine the accuracy of lesion type classification as defined by the Society of Cardiovascular Angiography and Interventions (SCAI) by MDCT using CA as the reference standard.

Methods

Patient population. This is a retrospective, single-center, observational cohort study. Data were collected from the Veterans Affairs Loma Linda Healthcare System (VALLHCS) MDCT database which included patients enrolled during a 21-month period from March 2005 to December 2006. Only patients who had undergone both MDCT and CA a maximum of 2 months apart were included. Patients who had undergone prior percutaneous or surgical revascularization were included as well.
MDCT protocol. MDCT was performed with a half-millimeter 16-detector-row scanner (Aquilion™ 16, Toshiba Medical Systems Corp., Otawara, Japan). The MDCT protocol has been previously described.5 Briefly, initial electrocardiographic (ECG)-triggered 3.0-mm-thick cross-sectional measurements were acquired for calcium score measurement prior to contrast infusion according to the original protocol validated by electron beam computed tomography.⁶ Subsequently, a 150 mL bolus of iodixanol was injected intravenously with a 50 mL saline chaser at a rate of 5 mL/second. Bolus tracking was performed by obtaining successive axial slices over the ascending aorta immediately after the beginning of radiocontrast infusion. As the signal in the ascending aorta reached a predefined threshold of 100 Hounsfield units (HU), the patient was instructed to sustain an inspiratory breath hold lasting about 20 seconds.
The scanning protocol used a gantry rotation time of 400 msec and slice collimation of 16 x 1.0 mm which were applied at a pitch of 3.2. The tube voltage was 135 kV and the current was 350 mAs. After acquisition of the raw data, retrospective ECG-synchronized slices were reconstructed by using a multisegment reconstruction algorithm.⁷ MDCT images were analyzed on a Vital Images Vitrea^® 2 workstation (Vital Images, Inc., Minneapolis, Minnesota) according to an established protocol.⁵ Initially, adjustments of different window and level settings are made on the three-dimensional volume renderings. Then, the best cardiac phase to analyze the left and right coronary systems is visually selected. At this point, it is possible to manually isolate the entire left and right coronary arterial systems with exclusion of other cardiac structures. This allows for a visually-guided application of the vessel probe on every artery, with emphasis on the origin, proximal and middle segments. Ultimately, this generates two orthogonal two-dimensional curved multiplanar reformation (CPR) images and a series of cross-sectional views. Finally, assessment of every orthogonal CPR and cross-sectional image is performed by the operator’s manipulation of the software.
Subsequently, in a blinded fashion, respective coronary angiograms were reviewed. Coronary lesion characteristics were categorized according to the SCAI lesion classification system⁸ (Figure 1):
• Type I lesion: patent, does not meet criteria for a type C lesion;
• Type II lesion: patent, meets criteria for a type C lesion;
• Type III lesion: occluded, does not meet criteria for a type C lesion;
• Type IV lesion: occluded, meets criteria for a type C lesion. A Type C lesion is one that is defined by one of the following criteria: • Diffuse (> 2 cm) length;
• Excessive tortuosities of proximal segments;
• Extremely angulated segments (usually > 90 degrees);
• Involvement of major side branches with inability to protect these branches;
• Degenerated vein grafts with friable lesions.

Statistical analysis. Analysis of MDCT (both two-dimensional and three-dimensional) and coronary angiographic images was limited to the origin, proximal and middle segments of the major coronary arteries and large branches (> 2 mm in diameter). These included the left main (LM), left anterior descending (LAD), circumflex (Cx), right coronary artery (RCA), large-caliber diagonal and obtuse marginal (OM) branches, ramus intermedius (RI) and left internal mammary artery grafts (LIMA). MDCT and angiographic images were electronically stored.
One lesion in each vessel segment comprised a data point for comparison between the two imaging modalities. Comparisons were then performed using the Spearman correlation test. The Spearman’s rank correlation coefficient was used to access the direction and strength of the relationship between the two nominal variables (lesion type by MDCT versus CA). The Statistical Package for the Social Sciences (SPSS 12.0 for Window; SPSS, Inc., Chicago, Illinois) was used for all analyses.

Results

A total of 411 segments were analyzed. Of this total amount, 110 (27%) were found to have lesions. Table 1 describes the lesion distribution by vessel. The Spearman correlation test revealed a statistically significant lesion type correlation between MDCT and coronary angiography for the LAD, Cx, RCA (p < 0.0001) and the OM (p = 0.0095). The sample size for the remaining vessel segments was too small to draw any meaningful statistical correlations, as shown in Table 2.
There were a total of 36 Type I, 39 Type II, 13 Type III, and 10 Type IV lesions detected by MDCT. Of the 49 patients found to have Types II and IV lesions, 45 (92%) of these patients were found to have Type C lesion characteristics due to lesion length. The mean lesion length in this subset of patients was 2.5 cm. The remaining 4 patients were found to have lesions involving major side branches.
Coronary arteriography detected a total of 16 Type I, 33 Type II, 14 Type III and 11 Type IV lesions. Of the 44 patients found to have Types II and IV lesions, 42 (95%) of these patients were found to have Type C lesion characteristics due to lesion length. The mean lesion length in this subset was 2.1 cm. The remaining 2 patients were found to have lesions jeopardizing major side branches.
Overall, there was a > 90% concordance when comparing high-grade (i.e., Types II, III and IV) lesions by MDCT versus CA. However, when comparing Type I lesions, the concordance was only 30% (Figure 2).

Discussion

Traditionally, invasive catheter-based CA has been used to image the coronary arteries, providing excellent temporal resolution and spatial resolution. In the past, noninvasive imaging modalities did not rival these parameters. However, improvements in MDCT technology now allow effective imaging of the epicardial coronary tree. Furthermore, this imaging modality now allows quantification of coronary lesion stenosis and identification of lesion characteristics.
Many studies have demonstrated that percent stenosis, as measured by MDCT, is comparable to percent stenosis determined by CA. Both 16- and 64-slice scanners demonstrate the sensitivity ranging from 80–90%, with specificity consistently being shown to be > 95%.^1–3
At present, no study has yet compared SCAI lesion characteristics between CA and MDCT. A qualitative assessment of SCAI lesion type may be very useful in planning coronary revascularization strategies. The limitation of CA is the fact that it represents a “lumenogram”. As such, qualification of lesions becomes challenging. For example, assessing vessel wall diameter versus lumen diameter can become difficult with CA. Furthermore, appreciation of the degree of calcification of lesions may be underestimated by CA and conventional fluoroscopy. MDCT can provide more insight into the differences between vessel wall diameter and lumen diameter, and can accurately measure lesion lengths. MDCT also very clearly identifies heavily calcified segments. All of these can in turn allow the operator to more effectively plan a percutaneous coronary intervention (PCI), from deciding on appropriate stent length and diameter, to considering the use of plaque modifying devices such as rotational atherectomy. This preplanning will likely allow the patient the benefit of less contrast and fluoroscopy exposure during PCI.
Our study suggests that there is a strong correlation between MDCT and CA when assessing high-grade lesions, i.e., SCAI Types II–IV. This may be of significant use in pre-procedural planning for PCI, as alluded to before. The weaker correlation observed with Type I lesions suggests that these types of lesions may need to be studied further when performing routine CA. Continued advancement of MDCT imaging technologies are likely to improve the accuracy of categorizing lesion types.
Study limitations. This was a single-center study utilizing a 16-detector scanner. With our current system, a 1 mm slice thickness was found to represent the best compromise between breath holds and resolution. Even though a majority of the studies were of good quality, we will likely see improvements with the 64-, 256- or the dual-source scanners due to shorter breath holds and improved slice thickness. The limited size of our study group did not allow for analysis of the diagonal, RI and LIMA. With increased numbers, we will be able to better compare these segments, as well as continuing to compare the major coronary vessels that were compared in this study.

Conclusion

MDCT provides an accurate method to categorize SCAI coronary lesion Types II–IV. In the future, MDCT may prove to be a very useful tool in planning coronary revascularization strategies.

1. Kuettner A, Trabold T, Schroeder S, et al. Noninvasive detection of coronary lesions using 16-detector multislice spiral computed tomography. J Am Coll Cardiol 2004;44:1230–1237.
   
2. Leber A, Knez A, von Ziegler F, et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography. J Am Coll Cardiol 2005;46:147–154.
   
3. Kuetter A, Kopp A, Schroeder S, et al. Diagnostic accuracy of multidetector computed tomography coronary angiography in patients with angiographically proven coronary artery disease. J Am Coll Cardiol 2004;43:831–839.
   
4. Achenbach S, Daniel W. Computed tomography of the coronary arteries. More than meets the (angiographic) eye. J Am Coll Cardiol 2005;46:155–157.
   
5. Cordeiro MAS, Lardo AC, Brito MSV, et al. CT angiography in highly calcified arteries: 2D manual vs. modified automated 3D approach to identify coronary stenosis. Int J Cariovasc Imag 2006;22:507–516.
   
6. Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15:827–832.
   
7. Dewey M, Laule M, Krug L, et al. Multisegment and halfscan reconstruction of 16-slice computed tomography for detection of coronary artery stenosis. Invest Radiol 2004;39:223–229.
   
8. Smith SC, Feldman TE, Hirshfeld JW, et al. Percutaneous coronary intervention: ACC/AHA/SCAI 2005 guideline update for (update of the 2001 PCI guidelines). J Am Coll Cardiol 2006;47:216–235.
   
9. Mercado N, Maier W, Boersma E, et al. Clinical and angiographic outcome of patients with mild coronary lesions treated with balloon angioplasty or coronary stenting. Implications for mechanical plaque sealing. Eur Heart J 2003;24:541–551.
   
10. Hecht HS, Roubin G. Usefulness of computed tomographic angiography guided percutaneous coronary intervention. Am J Cardiol 2007;99:871–875.
    
11. Deetjen AG, Conradi G, Mollmann S, et al. Diagnostic value of the 16-detector multislice computed tomography for the detection of coronary artery stenosis in comparison to invasive coronary angiography. Clin Cardiol 2007;30:118–123.