Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

CT Coronary Angiography


ACC/AHA 17-Segment Coronary Model

Comparative studies of coronary CT angiography (CTA) versus angiographic quantitative coronary analysis (QCA) assessment of coronary anatomy usually are performed comparing the presence or absence of angiographically significant stenosis (>50%) among the “17-segment model” of coronary anatomy.

Imaging Coronary Arteries: Luminal Assessment

Imaging of the coronary artery lumen is a particular challenge for CT scanning, given:

  • □

    The small luminal size (especially when diseased)

  • □

    The rapid motion (up to 150 mm/sec, particularly for the arteries in the atrioventricular grooves) through several phases of the cardiac cycle (early systole, early diastole and late diastole—if in sinus rhythm)

  • □

    The potential of rhythm or rate irregularity during scanning/acquisition

  • □

    The need for the patient to hold his or her breath during the procedure

  • □

    The presence of stents and the tendency of diseased coronary arteries to attract calcification—both of which confer problematic signal “blooming” (partial volume averaging) artifacts.

  • □

    The combination of motion, blooming artifact, and partial signal averaging effect that substantially reduces the depiction of the lumen within calcified and stented coronary arteries

Attaining equivalence in accuracy between CTA and catheter angiography is a challenge that has not been met as of 2014. Catheter angiography has substantially greater imaging resolution (spatial and especially temporal) and hugely greater versatility, yielding clinically adequate images despite calcification and stents, irregular cardiac rhythms, variable cardiac heart rates, and spontaneous or mechanical ventilation, on stable ambulatory patients and patients in cardiogenic shock ( Table 6-1 ).

TABLE 6-1
Catheter versus CT Angiography and MR Angiography
CATHETER ANGIOGRAPHY CT ANGIOGRAPHY MRI ANGIOGRAPHY
Resolution
Spatial (mm) 0.2 0.4–0.75 0.7–1.0 mm
Temporal (ms) 4–7 165–330
Dual source: 60–100		 0
3D rendering — Excellent Limited
Experience Huge Modest Limited
Radiation (mSv) 3–4 Newer scanners (mSv) Older scanners (mSv) None
Prospective 0.5–4 4–10
Retrospective 4–8 12–20
Ability to image/characterize the wall Poor (only dense and thick calcium is apparent) Good Poor

Thus, catheter-based angiography has logarithmically greater temporal resolution and several-fold better spatial resolution than contemporary CTA. CTA technology is developing in terms of both spatial and, more importantly, temporal resolution, but the technology gap is notable, and the clinical performance gap is broad in real life. Patients undergoing catheter-based coronary angiography are extremely heterogeneous with respect to many parameters that are relevant to CTA imaging:

  • □

    Variable heart rates:

    • Low (due to physiologic, pathologic or pharmacologic reasons) to tachycardic

    • β-blocker tolerant to β-blocker intolerant

    • Unstable heart rates

  • □

    Heart rhythm: normal sinus rhythm to arrhythmia of all forms

  • □

    Variable potential for artifact:

    • No calcification or minor calcification to heavy calcification

    • No stents

    • Multiple stents, including stenting within stenting

    • Stenting within calcified areas

Given the inherent lesser spatial resolution of CTA, its best coronary application, and probably its earliest, will be to establish that the lumens of large coronary vessels and bypass grafts are not stenotic.

Magnetic resonance angiography (MRA) is becoming increasingly capable and exhibits comparable sensitivity and specificity and accuracy to coronary CTA (at least the older 16-CT) when compared to QCA for vessels 1.5 mm or larger in very highly selected patients and select centers. Comparison of cardiac magnetic resonance (CMR) with current state-of-the-art equipment does not confirm contemporary equivalence of CMR and CTA. A 3-tesla system with increased signal-to-noise ratio (SNR) allows for increased spatial resolution in whole-heart MRA techniques.

Because the motion, size, and orientation of coronary arteries are different, the accuracy of 64-slice multislice CT (64-MSCT) for the detection of coronary stenoses depends on which vessel is being investigated, due to differences in vessel size and the amount of motion through diastole. The right coronary artery and the left circumflex artery are subject to more motion than the left anterior descending (LAD) artery due to the mechanical effect or motion imparted by atrial contraction ( Table 6-2 ).

TABLE 6-2
Accuracy of 64-Slice CT to Detect Coronary Lesions Compared with IVUS
From Leber AW et al. Quantification of obstructive and nonobstructive coronary lesions by 64-slice computed tomography: a comparative study with quantitative coronary angiography and intravascular ultrtasound. J Am Coll Cardiol. 2005;46(1):147.
SENSITIVITY (%) SPECIFICITY (%)
Left mainstem 100 100
Right coronary artery 83 100
Left anterior descending coronary artery 87 93
Left circumflex coronary artery 71 77
Total 84 91
IVUS, intravascular ultrasound.

In patients with normal heart rates, multisegment reconstruction algorithms tend to have superior diagnostic accuracy and image quality compared with half-scan reconstruction algorithms. Automatic selective phase acquisition software in large detector platforms (≥256 slice) and single cardiac cycle acquisition are expected to optimize acquisition with respect to avoidance of diastolic motion and image reconstruction artifacts resulting from multicycle reconstructions.

Calcium, Calcium Scoring, and Calcium Artifact

The presence of calcium, while useful for calcium scoring, is a problem for coronary CTA. The presence of calcium leads to false-positive determination of stenosis as well as overestimation of stenosis. It is more difficult to visualize the lumen in the presence of calcium, due to the “blooming artifact” (i.e., partial volume averaging) of calcium, which is responsible for overestimation of plaque size and stenosis severity. Any motion encountered during acquisition further compounds the blooming and increases the overestimation of stenosis and underestimation of lumen. Blooming artifact is seen less frequently with 64-CCT but remains a significant problem.

The addition of calcium scoring to coronary CTA does not add significantly to the determination of disease. A high calcium score should be seen as a reason not to proceed with CTA, however.

Evidence-Based Review of Coronary Cta

Since about 2005, an accelerating body of evidence has been developing that establishes that coronary CTA is feasible, but within limits that require careful patient selection and preparation. The many limits are represented by the numerous patient exclusions seen in published studies. As the technology has developed, the list of limitation, particularly in terms of patient selection and preparation, has diminished, but exclusions remain numerous and define the nature of use of CTA.

To date, CTA studies have been most useful in excluding disease or classifying disease, if it is present, rather than in performing the vital task of providing a roadmap to plan revascularization procedures in patients with significant disease.

The earliest series using 4-slice CCT equipment, entirely as expected, yielded results inferior to 16-CCT, which currently is considered the bare minimum technology to use for coronary CTA. Today the current standard is 64-CCT or >64-CCT, as presented in the tables in this chapter. Significantly, 64-CCT has not been proven equivalent to conventional angiography. The recent arrival of 256- and 320-CCT, which enable whole heart acquisition in a single cardiac cycle, is expected to provide further imaging benefits, and results of these techniques are eagerly anticipated. Whether they are not just improvements but improvements that render CCT comparable to conventional catheter angiography remains to be seen.

All studies published to date reflect a high degree of selection of patients, establishing that although CTA is feasible, it is not universally feasible, unlike catheter-based techniques. As the temporal resolution of coronary CTA improves, the possibility of including patients with heart rates greater than 75 bpm (who currently are excluded) is emerging, as is the possibility of including patients with irregular heart rhythms. As temporal resolution improves, false-negative CTA studies due to motion artifacts may become less common.

Most comparative studies performed before the introduction of 64-CCT in 2005 are notable for:

  • □

    Use of a segment-by-segment comparison, which predictably increases the power of the study

  • □

    Comparison of only vessels ≥1.5 (or ≥2.0) mm in luminal size versus all segments, which is not realistic in excluding CAD or in planning revascularization techniques

  • □

    Comparison of assessable vessels

More recent studies have strived to include:

  • □

    All segments, regardless of size

  • □

    Vessel comparison (single-, double-, triple-vessel disease)

  • □

    Patient comparison (disease vs. no disease)

To achieve the greatest plausibility and comparative power for conventional coronary angiography, the ideal and most credible studies of coronary CTA will entail no more exclusions than those of coronary angiography and will validate that the use of coronary CTA in the context of potential use of conventional angiography provides a significant outcome or cost benefit.

Summary of CTA for the Detection of CAD

The use of CTA for the detection of CAD is summarized in Figure 6-1 and Tables 6-3 through 6-10 .

Figure 6-1, A, Plot and table of per-segment sensitivity of multislice computed tomography–coronary angiography (MSCT-CA) compared with coronary angiography (CA). B, Plot and table of per-segment specificity of MSCT-CA compared with CA. CI, confidence interval; df, degrees of freedom.

TABLE 6-3
4-CCT Assessment of Native Coronaries
AUTHOR JOURNAL YEAR n NONASSESS. (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
Achenbach et al. Circ 2001 64 32 85 76 59 98
Becker et al. JCAT 2002 28 5 81 90 97 89
Knez et al. Am J Card 2001 42 6 78 98 — —
Kopp et al. Eur Heart J 2002 102 16 86 96 — —
Lau et al. Radiology 2005 50 — 79 95 — —
Nieman et al. Lancet 2001 31 27 56 97 — —
Nieman et al. Lancet 2001 31 27 81 97 — —
Nieman et al. Circ 2002 59 7 95 86 — —
Nieman et al. Heart 2002 53 30 58 76 — —
Nieman et al. Radiology 2003 24 33 85 73 — —
Vogl et al. Radiology 2002 64 — 75 91 — —
Kuettner et al. JACC 2004 66 — 37 99 — —
CCT, cardiac CT; nonassess., nonassessable; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-4
12-CCT Assessment of Native Coronaries
AUTHOR JOURNAL YEAR n NONASSESS. (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
Ropers et al. Circ 2003 77 12 93 92 79 97
CCT, cardiac CT; nonassess., nonassessable; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-5
16-CCT Assessment of Native Coronaries
AUTHOR JOURNAL YEAR n NONASSESS. (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
Achenbach et al. Eur Heart J 2005 57 5 94 97 — 99
Achenbach et al. Circ 2004 22 — 82 88 — —
Heuschmid et al. Am J Roent 2005 37 0 59 87 — —
CAC < 1000 93 94 68 99
Hoffmann M et al. JAMA 2005 103 6 95 98 — 99
Hoffmann U et al. Circ 2004 33 — 63 96 64 96
(good image quality) 82 93 — —
Kefer et al. JACC 2005 52 — 82 79 — —
Kuettner et al. JACC 2004 58 16 72 97 — —
(Agatston <1000) 98 98 — —
Kuettner et al. Heart 2005 72 7 85 98 — 96
Kuettner et al. JACC 2005 72 0 82 98 87 97
Martuscelli et al. Eur Heart J 2004 64 16 89 98 90 98
Mollet et al. JACC 2004 128 7 92 95 79 98
Mollet et al. JACC 2005 51 n/a 95 98 87 99
Morgan-Hughes et al. Heart 2005 58 2 83 97 — 97
Romeo et al. JACC 2005 53 12% transplant pt 83 95 71 95
Schuijf et al. Am J Cardiol 2005 45 6 83 97 — 97
Cademartiri et al. Radiol Med (Torino) 2005 60 2 93 97 99 86
Cademartiri et al. Am J Roentgenol 2006 38 0 92 96 87 97
Cademartiri et al. Radiol Med (Torino) 2005 40 0 96 96 86 99
Dewey et al. Invest Radiol 2005 129 9 83 86 — 96
Fine et al. Int J Cardiac Imaging 2004 50 2 87 97 — 98
Kaiser et al. Eur Heart J 2005 149 23 30 91 — 83
Aviram et al. Int J Cardiovasc Intervent 2005 22 — 86 98 — 98
Garcia et al. JAMA 2006 187 29 85 91 — 99
Gulati et al. Natl Med J India 2005 31 14 85 94 76 96
CCT, cardiac CT; nonassess., nonassessable; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-6
32-CCT
AUTHOR JOURNAL YEAR n NONASSESS. (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
Cordeiro et al. Heart 2005 30 20 76 94 — 96
CCT, cardiac CT; nonassess., nonassessable; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-7
40-CCT
AUTHOR JOURNAL YEAR n NONASSESS. (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
Lim et al. Clin Radiol 2005 30 0 99 98 94 99
CCT, cardiac CT; nonassess., nonassessable; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-8
64-CCT Assessment of Native Coronaries
AUTHOR JOURNAL YEAR n NONASSESS. (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
Leber et al. JACC 2005 59 64 97
No severe calcium 87 98
Leschka et al. Eur Heart J 2005 57 n/a 94 97 99
Mollet et al. Circ 2005 52 2 99 95 76 99
Raff et al. JACC 2005 70 12 86 95 66 98
Pugliese et al. Eur Radiol 2005 35 6 99 96 78 99
Ropers et al. Am J Cardiol 2006 84 4 93 97 100
Fine et al. Am J Cardiol 2006 66 6 95 96 97 92
Nikolaou et al. AJR 2006 72 6 97 79 96
Weustink et al. JACC Per segment 2007 100
1489
100		 95 [90–97] 95 [90–97] 95 [90–97] 95 [90–97]
Per patient 95 [90–97] 95 [90–97] 95 [90–97] 95 [90–97]
Schuijf et al. Am J Cardiol 2006 61 1 85 98
Ehara et al. Circ 2006 99 8 90 94
Pre-test probability Post-test negative Post-test positive
Meijboom et al. JACC 2007 254
105 High probability 87 17 96
83 Intermediate probability 53 0 88
66 Low probability 13 0 68
Shabestari et al. Am J Cardiol 2007 35 92 97 77 99
Agatston <100 83
Agatston >400 60
Meijboom et al. Heart 2007 104 92 91 60 99
Miller et al. 55 NEJM 2008 291 Agatston <600, >1.5 mm 85 [79–90] 90 [83–94] 91 [85–95] 83 [75–89]
Budoff et al. JACC 2008 230 1%
≥50% stenosis 55 95 83 64 99
≥70% stenosis 31 94 83 48 99
Meijboom et al. JACC 2008 360 Per patient 99 [98–100] 64 [55–73] 86 [82–90] 97 [94–100]
Per segment 88 [85–91] 90 [89–92] 47 [44–51] 99 [98–99]
CCT, cardiac CT; nonassess., nonassessable; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-9
64-CCT Assessment of Native Coronaries
Data from Schroeder S, Achenbach S, Bengel F, et al. Cardiac computed tomography: indications, applications, limitations, and training requirements: report of a Writing Group deployed by the Working Group Nuclear Cardiology and Cardiac CT of the European Society of Cardiology and the European Council of Nuclear Cardiology. Eur Heart J . 2008;29(4):531-556.
PATIENTS (NO.) NOT EVALUABLE (%) SENSITIVITY (%) SPECIFICITY (%) PPV (%) NPV (%)
701 3.8 (27/701)
(95% CI: 2.6–5.6)		 98 (398/404)
(95% CI: 95–99)		 90 (263/293)
(95% CI: 86–93)		 93 (394/4240)
(95% CI: 90–95)		 95 (263/273)
(95% CI: 93–98)
CCT, cardiac CT; NPV, negative predictive value; PPV, positive predictive value.

TABLE 6-10
64-CCT Assessment of Native Coronaries
Data from Hamon M, Biondi-Zoccai GG, Malagutti P, et al. Diagnostic performance of multislice spiral computed tomography of coronary arteries as compared with conventional invasive coronary angiography: a meta-analysis. J Am Coll Cardiol . 2006;48(9):1896-1910.
ANALYSIS NO. SENSITIVITY (95% CI) SPECIFICITY (95% CI)
Per segment 22,798 0.81 (0.72-0.89) 0.93 (0.90-0.97)
Per vessel 2,726 0.82 (0.80-0.85) 0.91 (0.90-0.92)
Per patient 1,570 0.96 (0.94-0.98) 0.74 (0.65-0.84)
CCT, cardiac CT.

Coronary CTA

Review of coronary CTA usually begins with a review of axial images and then moves on to maximum intensity projection (MIP) images and reformats (e.g., multiplanar reconstructions and cross-sectional multiplanar reconstructions) to follow the course of the coronary arteries. Workstation software has increasingly included features to assist with coronary (or other) vessel extraction and then depiction. “Seed” markers are placed by the origin of the vessel of interest and distally along its course. Using edge detection algorithms, the vessel is “extracted” and depicted. As the images of the extracted vessel become, to some extent, abstract (for example, a curved artery is depicted in a straightened fashion), reference MIP and 3D images often are displayed concurrently to depict the course of the vessel. These images are usually shown as curved reformations with a single-pixel-thick line coursing down the center of the vessel. Edge detection algorithms allow for maintenance of the center line positioning as the vessel curves. The curve can be represented in any degree of rotation around the center line, which can aid in visualizing eccentric lesions. In addition, cross-sectional “cuts” or planes perpendicular to the long axis of the curved reformations allow for circumferential evaluation of any segment of the analyzed coronary artery ( Figs. 6-2 through 6-23 ; , , , , , , , , , , , , , , , , , , , , ).

Figure 6-2, Multiple reconstructions from a cardiac CT study in a 58-year-old woman with a history of angina and an indeterminant myocardial perfusion imaging (MPI) study. The straight and curved reconstructions demonstrate two complex, severe lesions within the proximal right coronary artery (RCA). Cross-sectional evaluation across the more proximal lesion demonstrates mixed plaque, and near-total occlusion. The distal RCA has an approximately 50% soft plaque lesion proximal to the crux. A patent ductus arteriosus (PDA), however is free of disease and is a good-sized vessel. The patient went on to angiography, which demonstrated high-grade stenosis in the proximal RCA, and multiple proximal RCA collaterals. Faint contrast is seen coursing through the remainder of the RCA, but with poor visualization of the other lesions and the non-stenosed PDA. See Video 6-1 , Video 6-2

Figure 6-3, Same patient as Figure 6-2 . Multiple cardiac CT reconstructions through a large obtuse marginal branch in a 58-year-old woman with a history of angina and indeterminate myocardial perfusion imaging study. These images demonstrate mild proximal circumflex and obtuse marginal disease. The proximal portion of the large obtuse marginal branch has a severe 70% stenosis. Cross-sectional images demonstrate this to be a mixed plaque stenosis. Conventional coronary angiography (left anterior oblique projection) demonstrates a 70% proximal obtuse marginal branch stenosis. See Figures 6-1 and 6-2 and Video 6-1 , Video 6-2

Figure 6-4, A 57-year-old man with typical angina. The patient underwent a nuclear medicine stress perfusion study ( A and B ). Stress views are located superiorly, and the resting views are located inferiorly, with short-axis views on the left and long-axis views on the right. No evidence of a resting or stress perfusion defect is seen. The patient, however, subsequently was referred for a cardiac CT study. Images from that study follow ( C–E ), and corresponding conventional angiographic images are shown in parts F – H. C and F, There is a mixed 50% stenosis in the proximal left circumflex artery and a severe 70% stenosis in the mid-left circumflex artery, imaged on both the CCT and the coronary angiogram. D and G, A tight 80% to 90% mid-right coronary artery stenosis is present, imaged on both the CCT and the coronary angiogram. E and H, A complex 70% to 80% proximal-to-mid left anterior descending artery stenosis is present, imaged on both the CCT and the coronary angiogram. Severe triple-vessel disease was present. The lack of any corresponding regions of perfusion abnormality on the myocardial perfusion imaging study is plausibly due to the occurrence of balanced ischemia, a known challenge/pitfall of nuclear medicine imaging. See Video 6-3 , Video 6-4 , Video 6-5

Figure 6-5, A 57-year-old man with chest pain, and an equivocal myocardial perfusion imaging (MPI) study. A, A mild (<30%) stenosis within the mid-right coronary artery. B, MIP reconstruction of the left main coronary artery demonstrates a moderate amount of mixed plaque in the distal left main coronary artery causing a 50% to 60% left main stenosis. These findings were confirmed on the patient’s conventional angiogram to the right. The patient was referred for surgical consultation. See Video 6-6 , Video 6-7

Figure 6-6, A 74-year-old man presents with an acute coronary syndrome. A, Coronary CTA reveals a complete occlusion of the right coronary artery (RCA) due to mixed plaque that is mostly “soft.” B, Corresponding angiogram image corroborates the occlusion of the RCA. C, 3D volume-rendered image depicts the RCA occlusion and collaterals from the distal left anterior descending artery toward acute marginal territory. D, Maximum intensity projection image reveals the acute marginal branches by which (some of the) reconstitution of the ongoing RCA occurs. See Video 6-8.

Figure 6-7, Multiple images from a 47-year-old man with chest pain. A and B, Nuclear medicine myocardial perfusion imaging study demonstrated a moderate-sized, partially reversible inferior wall defect. Stress images are placed superiorly with rest images inferiorly. The patient went on to have coronary CT angiography ( C ). A second panel demonstrates a curved multiplanar reconstruction of the right coronary artery (RCA) ( D ). A severe soft plaque stenosis is seen in the mid-RCA. Stretched views of the RCA had been obtained with corresponding cross-sectional views in the proximal RCA, through the mid-RCA lesion, and just distal to the mid-RCA lesion. The cross-sectional views ( E and F ) confirm the severe stenosis and the soft plaque component to the stenosis. The bottom left images show a noncontrast calcium score study. The patient’s overall calcium score was zero. There is a moderate-sized area of low attenuation seen in the basal inferior wall. This is also identified on the axial source CTA image ( F ). This area of low attenuation represents fatty metaplasia, and a prior chronic infarct, corresponding with the nuclear medicine study. These CT images, however, are limited in their ability to determine how much residual viability is associated with the region of the prior infarct.

Figure 6-8, A 52-year-old man presented with typical angina and an indeterminant myocardial perfusion study. Curved reformatted images through the circumflex and obtuse marginal branch do not demonstrate any evidence of a circumflex system stenosis. The corresponding conventional angiogram confirms no obstructive circumflex lesion. See Figures 6-9 and 6-10 and Video 6-9 , Video 6-10 , Video 6-11 , Video 6-12

Figure 6-9, Same patient as Figure 6-8 . A maximum intensity projection image ( A ) and a curved multiplanar reformatted image ( B ) demonstrate a moderate to severe soft plaque lesion within the mid-left anterior descending artery (LAD). Cross-sectional evaluation of the lesion is seen in C and D . Quantitative analysis suggests an intermediate severity lesion at 60% to 70%. Selective angiography ( E ) of the left coronary artery confirms the 60% to 70% mid-LAD lesion. A fractional flow or reserve (FFR) evaluation across the lesion demonstrated an FFR of 0.90, not significant. See Figures 6-8 and 6-10 and Video 6-9 , Video 6-10 , Video 6-11 , Video 6-12

Figure 6-10, Same patient as Figure 6-8 . Reformatted multiplanar ( A and B ) and vessel extraction ( C ) images through the right coronary artery (RCA) demonstrate a severe soft plaque stenosis with subtotal occlusion of the distal RCA. The patient went on to have conventional angiography. Selective injection of the RCA demonstrated a corresponding severe stenosis of the distal RCA ( D ). The patient went on to have successful percutaneous transluminal coronary angioplasty of the lesion ( E ). See Figures 6-8 and 6-9 and Video 6-9 , Video 6-10 , Video 6-11 , Video 6-12

Figure 6-11, A 52-year-old man presented to the emergency department with chest pain, nonspecific ECG changes, and a mild increase in serum troponin levels. A and B, Curved multiplanar reconstruction images demonstrate an occlusion or high-grade stenosis in the proximal circumflex artery. The straightened view ( C ), with corresponding cross-sectional views ( D and E ) demonstrate partial and then no contrast within the lumen, more in keeping with an occlusion. The patient underwent emergent coronary angiography ( F and G ), confirming the finding of an occluded circumflex artery. He also has mild left anterior descending artery disease. See Video 6-13 , Video 6-14

Figure 6-12, Curved mutliplanar reconstructions and intravenous ultrasound (IVUS) views of the left anterior descending (LAD) artery demonstrate a severe lesion in the mid-LAD. On the curved multiplanar reconstructions ( A and B ) it appears to be 100%. By the IVUS views ( C ) there is a spot of contrast, indicating a small lumen. The angiographic view ( D ) reveals the lesion to be 95%.

Figure 6-13, A 62-year-old man with chest pain. His cardiac CT images demonstrate mild to moderate narrowing of the proximal left main coronary artery followed by moderate ectasia of the mid-to distal left main coronary artery. The distal left main coronary artery has a moderate amount of calcification. There is a severe ostial stenosis, which is primarily soft plaque in etiology, of the left anterior descending (LAD) artery that is 70% to 80% in severity. The corresponding conventional angiogram ( D ) demonstrates selective catheterization of the left coronary artery and confirms the findings seen on cardiac CT, with ectasia of the distal left main coronary artery, and a severe 80% ostial LAD stenosis. See Figure 6-14 and Video 6-15 , Video 6-16

Figure 6-14, Same patient as Figure 6-13 . This figure attempts to correlate cross-sectional views from a cardiac CT study with an intravenous ultrasound (IVUS) study performed at the time of conventional angiography. The paired IVUS and cross-sectional CT images are color-coded, and correspond with the points of acquisition marked on the curve multiplanar reformation at the top of the figure. A, Severe ostial LAD stenosis, which is almost completely soft plaque in etiology. B, Moderate calcium within the distal left main coronary artery, seen as peripheral curvilinear echogenicity with posterior shadowing on the IVUS image. C, Image obtained in the mid-left main coronary artery demonstrates mild eccentric soft plaque. D, Image obtained at the left main ostium/ascending aorta demonstrates no plaque. See Figure 6-13 and Video 6-15 , Video 6-16

Figure 6-15, A 68-year-old man presented with typical angina and shortness of breath on exertion. A nuclear medicine perfusion study ( A ) demonstrated a moderate-sized reversible defect involving the basal to mid-inferior wall. The patient was referred for CT angiography ( B ), which demonstrated proximal occlusion of the right coronary artery by an elongated complex lesion made up of soft and calcified plaque.

Figure 6-16, A 69-year-old man with a history of chest pain. A and B, CPR views of the left anterior descending artery (LAD) demonstrate a severe 70% ostial LAD stenosis. The mid-LAD lesion ( B ) represents misregistration artifact. A clean straight line is seen extending through the coronary artery, and into the adjacent soft tissues. C and D , Straightened view and corresponding cross-sections below it demonstrate mixed plaque extending from the distal left main coronary artery to the LAD. The linear demarcation on the straightened view ( C ) corresponds with the cross-sectional views ( D ). E and F, The patient went on to conventional angiography. Selective injection of the left coronary artery confirms the severe ostial LAD stenosis. See Video 6-17 , Video 6-18

Figure 6-17, A 55-year-old man presented to the emergency department with atypical chest pain and nonspecific ECG changes. CT angiographic maximum intensity projection and curved reformatted images ( A , B , C ) demonstrate multiple lesions within a large obtuse marginal branch. The proximal-most lesion is moderate, at 50%. The subsequent two lesions are severe, one at 80% and the other greater than 90%. The two severe lesions in the midportion of this obtuse marginal branch demonstrate an interposing segment of dilatation, and are constituted entirely by soft plaque. The patient was sent for coronary angiography. D demonstrates these two severe tandem lesions within the obtuse marginal branch, correlating well with the CTA. Angioplasty and stenting of the obtuse marginal branch were then carried out with a good result ( E ). See Figure 6-18 .

Figure 6-18, Same patient as Figure 6-17 . The CT angiographic maximum intensity projection ( B ) and curved reformatted images ( A ) demonstrate mild to moderate atherosclerotic ectasia of the proximal to mid-right coronary artery (RCA), but no evidence of a significant RCA stenosis. Severe OM1 stenoses were identified.

Figure 6-19, A 53-year-old man presented with atypical chest pain and multiple risk factors. A myocardial perfusion study demonstrated a possible mild anterior defect. Multiple reconstructed images through the right coronary artery (RCA) from a cardiac CT study demonstrate an unusual variant of RCA anatomy ( A – C ). The posterolateral branch arises from the proximal RCA and extends into the intraventricular groove. There is a severe (>70%) stenosis in the proximal portion of this posterolateral branch. The RCA demonstrates mild irregularity, but no evidence of a stenosis. These findings are confirmed on the left anterior oblique projection of the patient's conventional coronary angiogram ( D ). See Figures 6-20 and 6-21 and Video 6-19 , Video 6-20 , Video 6-21

Figure 6-20, Same patient as Figure 6-19 . A myocardial perfusion imaging study demonstrated a possible mild anterior defect. Multiple reconstructed images through the left anterior descending (LAD) artery demonstrate complex plaque within the proximal LAD. ( A – D ). At the level of the first diagonal branch takeoff there is a severe mixed plaque stenosis greater than 70%. The LAD is a large type III vessel ( A ), coursing well around the apex of the left ventricle. The severe proximal to mid-LAD stenosis is confirmed on the conventional angiogram ( E ). See Figures 6-19 and 6-21 and Video 6-19 , Video 6-20 , Video 6-21

Figure 6-21, Same patient as Figure 6-19 . A myocardial perfusion imaging study demonstrated a possible mild anterior defect. Multiple reconstructed images through the proximal circumflex artery demonstrate mixed plaque in the ostial/proximal circumflex artery causing a 50% stenosis. The corresponding conventional angiogram at the bottom of this figure confirms a 50% ostial circumflex artery stenosis. See Figures 6-19 and 6-20 and Video 6-19 , Video 6-20 , Video 6-21

Figure 6-22, CTA ( A ) and matching conventional angiography ( B ) of a patient with a severe stenosis of the proximal right coronal artery.

Figure 6-23, A through D, Contrast-enhanced cardiac CT curved multiplanar reformatted and intravenous ultrasound images demonstrating occlusion of the mid-right coronary artery due to mixed plaque. E through F, Myocardial perfusion scanning revealing an associated reversible inferior defect.

Comparison of Coronary CTA and QCA for Quantification of Stenosis

To date, only a few studies have published direct comparisons of degree of luminal narrowing by coronary CTA versus catheter angiography QCA or intravenous ultrasound (IVUS). The following observations are common to these studies:

  • □

    Coronary CTA correlates (imperfectly) with catheter-based angiographic determination of luminal narrowing.

  • □

    Limitations of spatial resolution, calcium, and blurring are seen.

  • □

    Coronary CTA is not able to accurately offer “percent stenosis” per case that agrees closely with catheter measurements.

  • □

    Older CCT coronary CTA systems tend to systematically overestimate luminal narrowing, especially because of their lower spatial and temporal resolution. This may be compounded by the greater blurring/blooming artifact from calcium. Newer systems exhibit less of this tendency due to improvements in spatial resolution and less of a tendency to exhibit blurring/blooming.

  • □

    The presence of calcium tends to increase the “positivity” of stenosis on CTA (it may confer false positives and overestimate coronary CTA assessment of severity, ) and may “overcall” complete occlusions.

  • □

    Coronary CTA may systematically underestimate percent area narrowing compared with IVUS.

  • □

    There is no correlation between stenosis difference and stenosis severity.

  • □

    Lack of temporal resolution seems to be responsible for some cases in which coronary CTA is not sensitive to angiographic stenosis.

  • □

    Use of segment comparison versus vessel comparison (one-, two-, or three-vessel disease) or patient comparison (diseased or not) provides an increase of the data that improves the significance of correlations but that does translate on a per patient basis.

  • □

    Small vessels tend to be underrepresented due to partial volume averaging effects and their common exclusion from studies.

The use of IVUS as a standard is understandable, but IVUS is not without imaging challenges. Determining the inner margin of the stenosis is usually straightforward, but determination of the outer margin of a plaque may be very difficult if the plaque is thick or dense with calcium. Establishing the outer elastic lamina by IVUS so as to determine plaque area or volume becomes increasingly difficult as the artery is increasingly diseased with plaque. Hence, some of the lack of correlation of CTA observations with IVUS is intrinsic to the limitations of CTA, but some is intrinsic to the limitations of IVUS ( Fig. 6-24 ). In one case, an unusual cause of coronary stenosis—a surgical suture—was detected by CCT ( Fig. 6-25 ).

Figure 6-24, A and B, 64-slice CT angiography (CTA). A, The diameter stenosis by multidetector CT (MDCT) correlates with those obtained by catheter-based quantitative coronary analysis (QCA) assessment ( r = 0.54), but tend to be 20% less. B, CTA plaque versus intravenous ultrasound (IVUS) plaque ( r = 0.61). CTA estimates the plaque area to be, on average, 9% less than by IVUS. C and D, 16-slice CTA. Quantitative coronary CTA correlates with QCA (0.75; P < .001), and significantly improves the diagnostic accuracy (receiver operator characteristic area under the curve [AUC] 0.81 versus 0.92; P < .001). E and F, 64-Slice CTA. Plaque volumes by MDCT correlate with those by catheter-based QCA assessment ( r = 0.76; P < .0001), but coronary CTA systematically underestimates volume. Lower right plot: Bland-Altman analysis, where the solid line indicates systematic error (the mean difference is 1.3 ± 14.2%) and the hatched lines indicate 95% agreement. There is no correlation between stenosis difference and stenosis severity. 92% of observations are within 25% error. CSA, cross-sectional area; DS MDCT, diameter stenosis MDCT; DS QCA, diameter stenosis quantitative coronary angiography; EEM, external elastic lamina; MSCT, multislice computed tomography.

Figure 6-25, A through C, Presurgical coronary angiogram that showed the absence of coronary artery stenoses. A, Right coronary artery. B, Left anterior descending coronary artery. C, Left circumflex coronary artery. D and E, Stenosis of the right coronary artery in coronary CT angiography performed on the fifth postoperative day. D, Multiplanar curved reconstruction of the right coronary artery in a coronary CT angiogram shows a high-grade, eccentric coronary artery stenosis (arrow) proximal to the right ventricular branch. E, Three-dimensional visualization of the right coronary artery stenosis ( large arrow ). Atrial temporary epicardial pacing lead (A); ventricular temporary epicardial pacing lead (V); pericardial drain (D). F and G: Invasive coronary angiography and percutaneous intervention performed after the CT angiogram. The right coronary artery stenosis is confirmed ( F, arrow ) and successfully treated by stent implantation ( G ). The suture had been placed to close an atrial cannulation site.

Reporting of Coronary Artery Stenosis Severity

Numerous reporting schemes have been employed to convey the severity of coronary artery stenosis.

  • □

    Visual estimate of the percent of stenosis:

    • The lack of highly reliable correlation with conventional angiography percent stenosis estimates dampens enthusiasm for reporting by percent stenosis.

    • However, supporters advocate use of percent stenosis on the basis of anticipation of better correlation with newer scanners, of intention to achieve good correlation, and of recognition of the clinical implications of potential for miscategorization if categories are used.

  • □

    There are variations in quartile description of stenoses ( Table 6-11 ).

    TABLE 6-11
    Quartile Descriptors of Stenosis Severity
    BASIC QUARTILES (%) ADAPTED QUARTILES (%) OTHER (%) 2008 KEY DATA ELEMENTS AND DEFINITIONS FOR CARDIAC IMAGING REPORT (%)
    <25

    26–50
    51–75
    76–100

    		<25

    26–50
    51–70
    71–100

    		Normal appearing: 0–24
    Mild: 25–49
    Moderate: 50–74
    Severe: ≥75     		Normal

    <50
    50–70
    >70–99
    Occluded

“Significant” is often defined as stenosis of more than 50% of the diameter and “severe” as stenosis of 70% to 75% of the diameter. This differs, however, from general coronary angiography, where “significant” is applied to lesions greater than 70%, except for the left main stem coronary artery, where “significant” means stenosis of more than 50%.

Issues with Catheter Estimates of Coronary Stenosis Severity

It is important to recall how imperfectly catheter-based angiography correlates with itself: interobserver laboratory variability: 10.4% ; interobserver variability: 11.2% ; and intraobserver correlation of visual assessment of catheter-based angiography and of QCA (percent stenosis: ±5%, minimal luminal diameter: ±0.15–0.28 mm). Furthermore, the visual angiographic estimate of coronary stenosis severity and QCA estimate of coronary stenosis severity do not necessarily correlate with fractional flow reserve determination of flow limitation, especially for lesions of moderate severity. Interobserver concordance tends to be poor (Spearman 0.36). For moderate lesions, visual estimate achieves good sensitivity (80%) and negative predictive value (91%) but poorer specificity (47%) and even less positive predictive value (25%) when compared with fractional flow reserve. “Frame bias” has been shown to significantly affect the variability of QCA coronary stenosis severity estimate in PCI. The use of different catheters (8F versus 6F) as the reference diameter confers variability to the QCA estimate of coronary stenosis severity (coefficient of variation 18.5% for minimal luminal diameter, 10.4% for percent stenosis). Contrary to expectation, experience in angiographic assessment does not necessarily continue to improve observer accuracy. Visual estimates may achieve better correlation than caliper estimates. Agreement tends to be better for proximal coronary segments than distal coronary segments and worse for moderate lesions.

Humblingly, there is poor correlation between the physiologic assessment of intermediate coronary stenoses (determined by fractional flow reserve) and the anatomic assessment by both visual and quantitative assessment of both coronary angiography and coronary CTA. The diagnostic accuracy in detecting a hemodynamically significant lesion (FFR < 0.75) for the different modalities and modes of interpretation is presented in Table 6-12 .

TABLE 6-12
Diagnostic Accuracy to Detect a Hemodynamically Significant Coronary Lesion by Modality
MODALITY/TECHNIQUE DIAGNOSTIC ACCURACY (%)
CTCA visual estimate 49
Quantitative CTCA 71
Coronary angiography—visual estimate 61
Quantitative coronary angiography 67
CTCA, CT coronary angiography.

There is poor correlation of the physiologic assessment of intermediate coronary stenoses (determined by fractional flow reserve) with anatomic assessment by both visual and quantitative assessment of both coronary angiography and coronary CTA. The diagnostic accuracies in detecting a hemodynamically significant lesion (FFR < 0.75) for the different modalities and modes of interpretation are as follows:

  • □

    Coronary angiography—visual estimate: 61%

  • □

    Coronary angiography—quantitative: 67%

  • □

    Coronary CT angiography—visual estimate: 49%

  • □

    Coronary CT angiography—quantitative: 71%

Ultimately, image analysis of coronary stenosis severity by catheter technique, CTA, or MRI inevitably entails degrees of variability ( Figs. 6-26 and 6-27 ). The variability of catheter-based determinations of coronary stenosis severity must be remembered when considering CTA correlation with catheter-based angiography used as the “gold” or reference standard.

Figure 6-26, Scatter plots of fractional flow or reserve (FFR) versus quantitative coronary angiography (QCA), quantitative computed tomography (QCT), conventional coronary angiography (CCA), and CT coronary angiography (CTCA). QCA, coronary angiography CCA, and CTCA are plotted versus FFR. There was a weak, but significant, negative correlation between QCA and FFR ( r = −0.30) and between QCT and FFR ( r = −0.32). Coronary arteries smaller than 3.5 mm are depicted as solid circles; coronary arteries larger than 3.5 mm are indicated as open circles.

Figure 6-27, Diagnostic performance of CT coronary angiography (CTCA) per segmental analysis categorized by diameter stenoses on quantitative coronary angiography (QCA). In the graph, the diagnostic performance of CTCA is shown according to various diameter stenoses as measured by QCA in a per-segment analysis. The absolute number of segments per stenosis category is shown in the table. The highest frequency of overestimated (FF) and underestimated (FN) coronary stenoses by CTCA was clustered around the cutoff value of 50% diameter reduction (significant coronary stenosis). FN, false negative; FP, false positive; TN, true negative; TP, true positive.

Characterization of Coronary Lesion and Arteries

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here

Related Posts