Author + information
- Received December 11, 2012
- Revision received April 2, 2013
- Accepted April 7, 2013
- Published online July 30, 2013.
- Ryo Nakazato, MD, PhD∗,
- Aryeh Shalev, MD∗,
- Joon-Hyung Doh, MD, PhD†,
- Bon-Kwon Koo, MD, PhD‡,
- Heidi Gransar, MS∗,
- Millie J. Gomez, MD§,
- Jonathon Leipsic, MD⋮,
- Hyung-Bok Park, MD∗,
- Daniel S. Berman, MD∗ and
- James K. Min, MD§∗ ()
- ∗Division of Nuclear Medicine, Department of Imaging, and Division of Cardiology, Department of Medicine, Cedars-Sinai Heart Institute, Cedars-Sinai Medical Center, Los Angeles, California
- †Department of Medicine, Inje University Ilsan-Paik Hospital, Goyang, Republic of Korea
- ‡Department of Medicine, Seoul National University Hospital, Seoul, Republic of Korea
- §Departments of Radiology and Medicine, Weill Cornell Medical College, Presbyterian Hospital, New York, New York
- ⋮Department of Medicine and Radiology, University of British Columbia, Vancouver, British Columbia, Canada
- ↵∗Reprint requests and correspondence:
Dr. James K. Min, Weill Cornell Medical College, New York Presbyterian Hospital, 520 East 70th Street, Starr Pavilion 8A, New York, New York 10021.
Objectives This study examined the performance of percent aggregate plaque volume (%APV), which represents cumulative plaque volume as a function of total vessel volume, by coronary computed tomography angiography (CTA) for identification of ischemic lesions of intermediate stenosis severity.
Background Coronary lesions of intermediate stenosis demonstrate significant rates of ischemia. Coronary CTA enables quantification of luminal narrowing and %APV.
Methods We identified 58 patients with intermediate lesions (30% to 69% diameter stenosis) who underwent invasive angiography and fractional flow reserve. Coronary CTA measures included diameter stenosis, area stenosis, minimal lumen diameter (MLD), minimal lumen area (MLA) and %APV. %APV was defined as the sum of plaque volume divided by the sum of vessel volume from the ostium to the distal portion of the lesion. Fractional flow reserve ≤0.80 was considered diagnostic of lesion-specific ischemia. Area under the receiver operating characteristic curve and net reclassification improvement (NRI) were also evaluated.
Results Twenty-two of 58 lesions (38%) caused ischemia. Compared with nonischemic lesions, ischemic lesions had smaller MLD (1.3 vs. 1.7 mm, p = 0.01), smaller MLA (2.5 vs. 3.8 mm2, p = 0.01), and greater %APV (48.9% vs. 39.3%, p < 0.0001). Area under the receiver operating characteristic curve was highest for %APV (0.85) compared with diameter stenosis (0.68), area stenosis (0.66), MLD (0.75), or MLA (0.78). Addition of %APV to other measures showed significant reclassification over diameter stenosis (NRI 0.77, p < 0.001), area stenosis (NRI 0.63, p = 0.002), MLD (NRI 0.62, p = 0.001), and MLA (NRI 0.43, p = 0.01).
Conclusions Compared with diameter stenosis, area stenosis, MLD, and MLA, %APV by coronary CTA improves identification, discrimination, and reclassification of ischemic lesions of intermediate stenosis severity.
- coronary artery disease
- coronary computed tomography angiography
- coronary plaque
- fractional flow reserve
- myocardial ischemia
Invasive coronary angiography (ICA) is an established technique for the assessment of coronary arteries in patients with suspected or known coronary artery disease (CAD) that provides useful anatomic information about the degree of luminal narrowing. Mild luminal narrowing of lesions of intermediate angiographic severity on ICA may serve as a surrogate for early-stage coronary atherosclerosis. This hypothesis is strengthened by the findings that diffuse mild angiographic luminal narrowing on ICA is associated with reduced stress-induced myocardial blood flow and abnormal resistance of the epicardial coronary arteries even before a high-grade segmental stenosis is apparent (1–3).
In part because of these findings, high-grade angiographic luminal narrowing visualized at the time of ICA does not reliably determine the hemodynamic significance of coronary lesions. In patients with intermediate coronary stenosis, clinical decision making often relies on supplemental functional testing (4). Fractional flow reserve (FFR) is a lesion-specific technique to determine the hemodynamic significance of a coronary stenosis and is widely considered the gold standard for physiological assessment of coronary stenoses.
Coronary computed tomography angiography (CTA) is a noninvasive test that correlates favorably with ICA and intravascular ultrasound for measures of the severity of luminal stenosis and direct visualization of coronary atherosclerosis, respectively (5). Methods of measuring the severity of luminal stenosis by coronary CTA, similar to ICA, include percent diameter stenosis, percent area stenosis, minimal lumen diameter (MLD), and minimal lumen area (MLA). Measures of coronary atherosclerosis by coronary CTA include aggregate plaque volume.
Although the extent of mild luminal angiographic narrowing is associated with the increasing prevalence and severity of global myocardial ischemia on myocardial perfusion imaging, the degree to which measures of coronary atherosclerosis contribute to this phenomenon is unknown. Further, the association of coronary atherosclerosis and ischemia in a lesion-specific fashion has not been evaluated to date. To address these issues, we examined coronary CTA measures of luminal arterial narrowing as well as cumulative coronary atherosclerotic plaque burden to identify the optimal methods for diagnosis of ischemia-causing lesions of intermediate stenosis severity when compared with an invasive FFR reference standard.
Fifty-eight consecutive patients from 2 centers who underwent clinically indicated coronary CTA and were visually observed to have a coronary CTA–determined isolated intermediate diameter stenosis (30% to 69%) in the proximal or mid-portion of a major epicardial coronary artery were studied. Upon visual identification of the intermediate stenosis, quantitative coronary CTA assessment was performed as described in the following text. Study patients underwent ICA with intended FFR of the intermediate stenosis within 3 months of the coronary CTA. Exclusion criteria included additional intermediate- or high-grade stenoses within the same vessel. The institutional review board of the 2 study centers approved the study, and all patients provided written informed consent.
Acquisition of coronary CTA
Coronary CTA was performed using dual-source computed tomography (CT) (Somatom Definition, Siemens, Forchheim, Germany) or 320-detector row CT (Aquilion One, Toshiba, Otawara, Japan) in direct accordance with the Society of Cardiovascular Computed Tomography guidelines on performance of coronary CTA (6). All patients were treated with nitroglycerin, and those with a heart rate >65 beats/min were treated with beta-blockers unless contraindicated. An intravenous bolus (60 to 90 ml) of contrast agent (Omnipaque 350 mg/dl or Visipaque 320 mg/dl; both from GE Healthcare, Princeton, New Jersey) was injected at a flow rate of 6 ml/s. The scan parameters included 64 × 0.625/0.750-mm collimation, a tube current time product of 350 mAs per rotation, and a tube voltage of 100 or 120 kVp. Scans were performed using electrocardiogram-based tube current modulation whenever possible or sequential (prospective) image acquisition. Transaxial images were reconstructed with 0.6-/0.75-mm slice thickness, 0.3- to 0.4-mm slice increment, and a medium-smooth convolution kernel. For the helical scan, the position of the reconstruction window within the cardiac cycle was individually selected to minimize artifacts. Motion-free datasets, typically in mid-diastole, were collected for analysis. The estimated effective radiation dose for coronary CTA ranged from 2 to 10 mSv using the dose-length product with an organ-specific conversion factor k of 0.014 mSv/mGy/cm (7).
ICA and FFR
Selective ICA was performed by standard catheterization in accordance with the American College of Cardiology guidelines for coronary angiography (8). FFR was measured using a 0.014-inch pressure sensor tipped guidewire (PressureWire, St. Jude Medical Systems, St. Paul, Minnesota) as previously described (9). Hyperemia was induced with intracoronary bolus administration (80 μg in the left coronary artery, 40 μg in the right coronary artery) or intravenous continuous infusion (140 μg/kg/min) of adenosine (10). Intracoronary nitroglycerin was administered immediately before measurement of FFR. In keeping with prior multicenter studies, an FFR value ≤0.80 was considered to indicate lesion-specific ischemia (11).
Coronary CTA data analysis
Coronary CTA was analyzed in blinded fashion and manually performed by 2 Level III experienced readers in accordance with the Society of Cardiovascular Computed Tomography guidelines on interpretation of coronary CTA (12). All studies were of good or excellent quality and allowed for comprehensive analysis of the coronary artery and atherosclerotic plaque features. Study results were computed as the average of 2 readers. The technique for coronary CTA measures was as follows. After centerline definition, contiguous cross-sectional reconstructions were rendered using a slice thickness of 1 mm on a dedicated 3-dimensional workstation (AW Advantage, GE Healthcare) (13,14). To allow for optimal and consistent detection of plaque and outer vessel boundaries, the window width was set as 155% of the mean intensity within the lumen and the window level was set at 65% of the mean intensity for each lesion, as previously described (14,15). Diameter stenosis (%) and area stenosis (%) were calculated using proximal and distal reference segments, which were selected to be the most adjacent points to the maximal stenosis in which there was minimal or no plaque. MLD (mm) and MLA (mm2) were measured from the long-axis and short-axis views of double-oblique reconstructions at the site of the maximal stenosis, respectively.
Percent aggregate plaque volume was calculated as follows. Outer vessel contours were identified and defined as the visualized border at which point low attenuation (epicardial fat) was observed; lumen areas were manually traced in each cross section. Plaque area was computed as the vessel area minus the lumen area of each cross section. Aggregate plaque volume was obtained by summation of all contiguous plaque areas from the coronary artery ostium to the distal portion of the lesion. Percent aggregate plaque volume was calculated as the aggregate plaque volume divided by the total vessel volume and reported as percentage. Figure 1 depicts a sample illustration of plaque measurements.
Continuous variables are expressed as mean ± SD. For comparison of 2 groups, Student t test or Wilcoxon rank sum test was used as appropriate. Categorical variables were compared using Pearson chi-square test or Fisher exact test as appropriate. Correlations between coronary CTA parameters and FFR were assessed by calculating Pearson correlation coefficient. Interobserver and intraobserver variability of coronary CTA parameters was assessed by intraclass correlation coefficients (ICC) for absolute agreement of single measures between consistent raters (MedCalc version 12.4, MedCalc Software, Ostend, Belgium) and Bland-Altman plots. Univariable logistic regression analysis was performed to evaluate the relationship between coronary CTA parameters and the presence of lesion-specific ischemia, where standardized odds ratio (OR) coefficients were added to enable direct comparison of predictors. For the multivariable logistic regression analysis, ORs were reported to examine the percent aggregate plaque volume versus diameter stenosis, area stenosis, MLA, and MLD individually.
To examine discrimination, area under the receiver operating characteristic curves (AUC) were constructed and compared for different coronary CTA parameters using the method described by DeLong et al. (16) for correlated data. We determined the optimal method of ischemia diagnosis and discrimination and evaluated the net reclassification improvement (NRI) of this method compared with other coronary CTA parameters (17). To define categories of risk of lesion-specific ischemia, we utilized anatomic measures of luminal stenosis severity to categorize lesions conventionally believed to be of low, intermediate, and high probability of ischemia, defined as <50%, 50% to 69%, and ≥70% for diameter stenosis and area stenosis. For MLD, MLA, and percent aggregate plaque volume, we used tertiles of these categories to examine reclassification of lesion-specific ischemia. For NRI calculations, cases were defined as patients with lesion-specific ischemia (FFR ≤0.80) and controls were defined as those without lesion-specific ischemia (FFR >0.80). Associations and differences with p values <0.05 were considered significant. Statistical analyses were performed using Stata software (version 11, StataCorp LP, College Station, Texas) except where otherwise noted.
We identified 58 intermediate stenoses in 58 patients. The median intertest duration between coronary CTA and FFR was 21 days (interquartile range: 13 to 35 days), with no adverse clinical events, change of symptoms, or coronary revascularization occurring between coronary CTA and FFR. The baseline characteristics of the study population are listed in Table 1. Intermediate lesions were most often present in the left anterior descending artery (n = 37), followed by the right coronary artery (n = 16) and then the left circumflex artery (n = 5). Twenty-two ischemic lesions (38%) were identified by FFR, with an average FFR value of 0.82 ± 0.10. Compared with nonischemic lesions, ischemic lesions demonstrated no differences in diameter stenosis or area stenosis but had smaller MLD and MLA and greater percent aggregate plaque volume (Table 2).
Correlation between coronary CTA parameters and FFR
Moderate significant correlations were observed between FFR values and diameter stenosis (r = −0.35, p = 0.007), area stenosis (r = −0.39, p = 0.003), MLD (r = 0.43, p = 0.0009), MLA (r = 0.47, p = 0.0002), and percent aggregate plaque volume (r = −0.50, p <0.0001). No correlation of percent aggregate plaque volume and lesion length was observed (r = −0.10, p = 0.4).
Interobserver and intraobserver variability
Interobserver variability for all lesions for measures of luminal narrowing and atherosclerotic plaque was good to excellent with generally low limits of agreement as follows: percent aggregate plaque volume: ICC 0.56, p < 0.0001, Bland-Altman bias 3.8%, limits of agreement −11.7% to 19.2%; diameter stenosis: ICC 0.68, p < 0.0001, Bland-Altman bias −2.6%, limits of agreement −25.3% to 20.1%; area stenosis: ICC 0.78, p < 0.0001, Bland-Altman bias −4.1%, limits of agreement −28.9% to 20.7%.
Intraobserver variability for a randomly selected 20% of lesions measured more than 60 days apart demonstrated excellent correlation with low limits of agreement as follows: percent aggregate plaque volume: ICC 0.99, p < 0.0001, Bland-Altman bias −0.4%, limits of agreement −3.0% to 2.3%; diameter stenosis: ICC 0.94, p < 0.0001, Bland-Altman bias −1.9%, limits of agreement −12.3% to 8.5%; area stenosis: ICC 0.97, p < 0.0001, Bland-Altman bias −0.5%, limits of agreement −13.0% to 11.9% (Online Fig. 1).
Diagnosis of lesion-specific ischemia by coronary CTA
On univariable logistic regression analysis, MLD, MLA, and percent aggregate plaque volume were significantly related to ischemia-causing intermediate lesions (Table 3). On multivariable logistic regression analysis, percent aggregate plaque volume remained a significant predictor for lesion-specific ischemia over diameter stenosis (OR: 1.28; 95% confidence interval [CI]: 1.12 to 1.47; p < 0.001), area stenosis (OR: 1.29; 95% CI: 1.12 to 1.49; p < 0.001), MLD (OR: 1.26; 95% CI: 1.11 to 1.43; p < 0.001), and MLA (OR: 1.27; 95% CI: 1.10 to 1.47; p < 0.001). Study results were unchanged when examined by particular CT system.
Discrimination of ischemia of intermediate stenoses by coronary CTA
The AUC for lesion-specific ischemia was higher for percent aggregate plaque volume (0.85; 95% CI: 0.74 to 0.97) than for diameter stenosis (0.68; 95% CI: 0.54 to 0.83), area stenosis (0.66; 95% CI: 0.51 to 0.80), MLD (0.75; 95% CI: 0.61 to 0.88), or MLA (0.78; 95% CI: 0.64 to 0.91) (Fig. 2). On adjusted logistic regression analyses, percent aggregate plaque volume provided incremental prediction for lesion ischemia over diameter stenosis (AUC 0.88 [95% CI: 0.78 to 0.99] vs. 0.68 [95% CI: 0.54 to 0.83], respectively; p = 0.02), area stenosis (0.88 [95% CI: 0.79 to 0.97] vs. 0.66 [95% CI: 0.51 to 0.80], respectively; p = 0.003), and MLD (0.90 [95% CI: 0.79 to 0.99] vs. 0.75 [95% CI: 0.61 to 0.88], respectively; p = 0.03). Percent aggregate plaque volume trended toward but did not significantly improve the discriminatory power over MLA (0.89 [95% CI: 0.79 to 0.99] vs. 0.78 [95% CI: 0.64 to 0.91], respectively; p = 0.07).
NRI of ischemic lesions
Percent aggregate plaque volume enabled effective reclassification of intermediate ischemic lesions with measures of diameter stenosis (NRI 0.77; 95% CI: 0.41 to 1.12; p < 0.001), area stenosis (NRI 0.63; 95% CI: 0.27 to 0.99; p = 0.002), MLD (NRI 0.62; 95% CI: 0.30 to 0.93; p = 0.001), and MLA (NRI 0.43; 95% CI: 0.11 to 0.75; p = 0.01) (Online Table 1). Figures 3 and 4 show an example of coronary CTA measures for an ischemic intermediate lesion and a nonischemic intermediate lesion, respectively.
In this study, we compared an array of CAD severity measures by coronary CTA for the diagnosis of ischemia for stenoses of intermediate severity. These quantitative measures included diameter stenosis, area stenosis, MLD, and MLA. Given the ability of coronary CTA to observe both luminal narrowing and atherosclerotic plaque, we also evaluated aggregate plaque volume, which considers the cumulative atherosclerotic plaque burden from the vessel ostium to the distal portion of a plaque causing a stenosis. We observed that percent aggregate plaque volume allowed for diagnosis of lesion-specific ischemia of intermediate stenoses in a manner that surpasses more traditional angiographic measures of CAD lesion severity and that, when combined with these measures, offered independent and incremental discriminatory power. The robustness of percent aggregate plaque volume as an important marker of ischemia was evident for all analyses, including diagnosis, correlation, discrimination, and reclassification.
Prior reports have consistently demonstrated that visual assessment of lumen diameter stenosis, the most commonly used method for determination of CAD severity, correlates unreliably to the presence of hemodynamically significant stenoses. These findings have evoked concerns that reliance on visualized measures of anatomic luminal narrowing alone may encourage coronary revascularization of nonischemia-causing coronary lesions. In contrast, non-negligible rates of ischemia across coronary lesions not traditionally considered angiographically severe have raised concerns that symptomatic patients with ischemia may be missed. These findings hold true for stenoses visualized by both ICA as well as coronary CTA and, in particular, for lesions of intermediate severity even determined by quantitative measures (1,18–20).
The potential explanations for these findings are manifold. Numerous additional factors may contribute to the induction of ischemia across a coronary stenosis and include anatomic factors, such as stenosis length or geometry, and physiological factors, such as impaired hyperemic responses arising from endothelial dysfunction or dynamic vasoconstriction. Directly germane to the present study, diffuse atherosclerotic plaque proximal to and involving a stenosis may also substantively contribute to the ischemia-producing nature of a coronary lesion, a finding that may result in the underestimation of diameter stenosis and area stenosis severity, in particular for ICA, by erroneous comparison of diseased segments to angiographically normal-appearing segments that nevertheless are affected by atherosclerotic plaque (21).
Pathophysiologically, the totality of atherosclerosis proximal to a coronary lesion has been demonstrated to be vital to the contribution of ischemia. De Bruyne et al. (3) reported that early-stage coronary atherosclerosis is often associated with abnormal resistance of the epicardial coronary arteries before a high-grade segmental stenosis is apparent at angiography by ICA. In addition to the resistance caused by focal stenosis or arteriolar vasomotor dysfunction, diffusely atherosclerotic epicardial coronary arteries without high-grade segmental stenoses often manifest a continuous pressure decline along their length, reduce coronary flow reserve, and contribute to myocardial ischemia. These findings suggest the importance of measuring the atherosclerotic plaque in stenotic lesions as well as angiographically “normal” lesions that may include diffuse atherosclerosis.
To date, several prior efforts have attempted to examine the ability of coronary CTA measures to identify lesion-specific ischemia. In a recent study by Kristensen et al. (22) of patients with intermediate stenoses undergoing evaluation with both ICA and coronary CTA, traditional quantitative coronary CTA parameters were better correlated to hemodynamic significance as compared with FFR. Our findings are in general accordance with this study, and we observed similar correlations between coronary CTA parameters and FFR as compared with these investigations. Nevertheless, our study directly builds on these important results by inclusion not only of luminal narrowing measures but the totality of atherosclerotic plaque volume proximal to the narrowing as well. We observed that diffuse atherosclerosis that was mild in luminal narrowing contributes significantly to the ischemic process. Further, we observed that the cumulative summation of atherosclerotic plaque before a lesion not only performs superiorly to measures of luminal narrowing but is also additive to luminal measures as well. In this regard, combined use of percent aggregate plaque volume and measures of luminal narrowing may represent a more ideal approach to identifying lesions that cause ischemia.
In this study, we restricted analysis to lesions of intermediate stenosis severity as visualized by coronary CTA. Anatomic measurement of luminal narrowing is difficult for these types of lesions, both by invasive and noninvasive means, given that the conventional definitions of angiographic severity for these lesions are at a point above which considerable rates of ischemia are often identified. This issue is underscored by the almost 40% of intermediate lesions in the current study that were associated with ischemia, with 6 of 23 lesions (26%) with diameter stenosis <50% causing ischemia.
This study is not without limitations. This study was performed on a generally small sample. Given the safety concerns of performing invasive FFR for nonclinically indicated coronary lesions, we reserved analysis only for those for whom clinical indications such as persistent angina drove the prospective performance of invasive FFR. Further, we only examined stenoses of intermediate severity, given the difficulty of diagnosing ischemia for these lesions on the basis of angiographic standards alone. This may have amplified the incremental utility of percent aggregate plaque volume, which may be significantly less robust in lesions of higher or lower stenosis severity. Future studies examining the relationship of percent aggregate plaque volume to these types of coronary lesions should be performed and are an active area of investigation in our own laboratory. We also identified intermediate coronary stenoses by visual assessment of an expert reader at the time of coronary CTA interpretation. We did so to reflect a “real-world” design for practitioners who often identify symptomatic patients with stenoses that may not necessarily be considered anatomically severe. The clinical interpretations that drove the included study sample were accurate, because the average diameter stenosis was 50.8%, with 52 lesions (90%) ranging between 30% and 69% diameter stenosis on quantitative coronary CTA. Finally, manual plaque measurements by coronary CTA, including those required to calculate aggregate plaque volume, are time consuming and necessitate expert interpretation. Whether the present study results can be universally applied to practitioners interpreting coronary CTA on a daily basis requires further study. Because several automated methods for coronary plaque quantification have recently been introduced, these software applications may allow for greater widespread adoption of the current study findings.
For coronary lesions of intermediate stenosis severity, percent aggregate atherosclerotic plaque volume is superior as well as additive to traditional angiographic measures of luminal narrowing for diagnosis, correlation, discrimination, and reclassification of ischemia.
For a supplementary figure and table, please see the online version of this article.
Dr. Leipsic is a member of the Speakers’ Bureau for GE Healthcare. Dr. Min serves on the medical advisory board for GE Healthcare, Arineta, AstraZeneca, and Bristol-Myers Squibb; receives research support from GE Healthcare, Philips Healthcare, and Vital Images; is a member of the Speakers’ Bureau for GE Healthcare; is a consultant for AstraZeneca and Bristol-Myers Squibb; and has an equity interest in TC3 and MDDX. All other authors have reported that they have no relationships relevant to the content of this paper to disclose.
- Abbreviations and Acronyms
- area under the receiver operating characteristic curve
- coronary artery disease
- confidence interval
- computed tomography
- computed tomography angiography
- fractional flow reserve
- invasive coronary angiography
- intraclass correlation coefficients
- minimal lumen area
- minimal lumen diameter
- net reclassification improvement
- odds ratio
- Received December 11, 2012.
- Revision received April 2, 2013.
- Accepted April 7, 2013.
- American College of Cardiology Foundation
- Gould K.L.,
- Kelley K.O.,
- Bolson E.L.
- De Bruyne B.,
- Hersbach F.,
- Pijls N.H.,
- et al.
- Bartúnek J.,
- Sys S.U.,
- Heyndrickx G.R.,
- Pijls N.H.,
- De Bruyne B.
- Min J.,
- Shaw L.,
- Berman D.
- Scanlon P.J.,
- Faxon D.P.,
- Audet A.M.,
- et al.
- MacCarthy P.,
- Berger A.,
- Manoharan G.,
- et al.
- De Bruyne B.,
- Pijls N.H.,
- Barbato E.,
- et al.
- Nakazato R.,
- Shalev A.,
- Doh J.H.,
- et al.
- Leber A.W.,
- Becker A.,
- Knez A.,
- et al.
- Kern M.J.,
- Lerman A.,
- Bech J.W.,
- et al.
- Tobis J.,
- Azarbal B.,
- Slavin L.
- de Feyter P.J.,
- Serruys P.W.,
- Davies M.J.,
- Richardson P.,
- Lubsen J.,
- Oliver M.F.