Coronary artery calcification: comparison between triggered and non-triggered tests
JAVIER COURTIS, GUSTAVO PARISI, MANUEL VACA VILLARROEL, PAULA RIBA,
SERGIO LUCCINO, MAGDALENA DIMITROFF, ANALÍA GONZALEZ
Instituto Oulton (Córdoba, Argentina); Facultad de Cs. Exactas, Físicas y Naturales; Facultad de Biología, Universidad Nacional de Córdoba (Córdoba, Argentina).
(5000) Córdoba Capital, Argentina E-mail
Recibido 21-DIC-2018 - ACEPTADO después de revisión el 07-FEBRERO-2019.
There are no conflicts of interest to disclose.
Patients who undergo a computed tomography (CT) scan of the chest for COPD or lung cancer detection, usually have risk factors for coronary heart disease. Since CT clearly identifies coronary calcification, it can be advantageous to detect two pathologies with great clinical impact in the same examination. In the present study, we sought to determine the correlation between the "Agatston Method", traditionally used in the evaluation of coronary calcium score (CCS) through multidetector computed tomography (MDCT) with triggered ECG, and a "Visual Method" by Chest CT without ECG. Methods and Results: Fifty consecutive patients who underwent a chest CT scan without ECG and without contrast for chest evaluation, were recruited to undergo CCS. Dividing the coronary arterial tree into 18 segments, CCS was measured in the traditional way, through the "Agatston Method". In turn, in the "Visual Method" by CT without ECG, a point was assigned to each coronary artery segment that presented parietal calcification (defined as all coronary arterial tissue ≥130 HU and an area of lesion ≥1 mm2), being the total visual coronary calcium score (TVCCS) for each subject, the sum of all the scores obtained in the four main arteries. Results. The TVCCS was categorized as follows: absent (0), mild (1-3), moderate (4-6) and severe (≥7), intended to be consistent with Agatston scores of 0, 1-100, 101-400 and ≥401, respectively. Using the Spearman’s rank correlation coefficient, a statistically significant association was found between the "Agatston Method" and "Visual Method" variables, estimated value of correlation r = 0.97 (p<0.01). In addition, a positive correlation was observed between both methods with the Chi-square hypothesis test with an observed proportion of 0.92 (p<0.01). Conclusions. This study validates the reliability of the "Visual Method" by CT without ECG to determine the amount of coronary calcium and its correlation with the "Agatston Method". In addition, this technique should allow the stratification of the risk of coronary atherosclerotic disease in patients undergoing a chest CT scan during the evaluation of pulmonary, aortic and / or mediastinal pathologies without requiring additional scanning.
Of all diagnostic tests available currently, and proposed to identify asymptomatic individuals in risk of suffering (fatal or non-fatal) acute myocardial infarction in a long term, coronary calcium score (CCS) through the Agatston method  has emerged as the most accurate and efficient tool . Over time, CCS has turned into the noninvasive method of choice to visualize directly coronary artery calcification (CAC) ; besides having proven to be a strong predictor of congestive heart failure  and stroke in asymptomatic patients .
As yearly, numerous computerized axial tomography scans (CAT), of the ECG-triggered chest kind, are conducted for the diagnosis of pulmonary diseases  in patients with an advanced age and history of smoking, and considering the latter are also associated to atherosclerotic coronary artery disease (CAD), and also knowing that more than half of this population group will present some degree of CAC [7,8], and that CAC identified in ECG-triggered chest CT clearly indicates atherosclerosis [9,10]; all these arguments lead us to think that it would be very advantageous to have a single diagnostic method available, capable of simultaneously detecting atherosclerotic CAD and chest pathologies (lung cancer, emphysema, etc.) with very low ionizing radiation doses.
For all these reasons, and due to small previous studies [11,12] reporting that in more than 90% of patients who undergo ECG-triggered multidetector computed tomography (MDCT) to determine CCS, the values obtained would present a high correlation to those obtained by ECG-triggered chest CT, and as also currently there is no specific recommendation on how to report the degree of CAC in this type of chest CT, it was proposed to perform this test with the aim of determining the correlation existing between the Agatston method estimated from the images obtained in MDCT and a method based on the “visual” observation of the images obtained from ECG-triggered chest CT.
All individuals with no history of known CAD and referred to our center for ECG-triggered chest CT for imaging tests due to general chest pathologies, were invited to participate in our study. Demographic, clinical, anthropometric and lab data, and information on any type of medication used were collected and clearly detailed at the time of the test by specific questionnaires. All patients that participated of this study gave their written consent, and those that refused to sign it were excluded. The signed informed consent was given before making both CT scans. Both the study protocol and the informed consent were evaluated and authorized by the Institutional Committee on Health Investigation Ethics (CIEIS Oulton).
Two experts in advanced cardiac imaging (JC and GP, 12 years and 5 years of experience in cardiac CT respectively) read and quantified CAC through both methods (Agatston and visual). Both acquisitions (triggered and non-triggered) were obtained with patients in supine decubitus position, with no angulation, and with a single breath held for 10 to 15 seconds approximately, minimizing artifacts by respiratory movement. Coronary artery calcium was defined (same definition for both methods) as any vascular tissue with an attenuation coefficient greater than 130 Hounsfield units (HU), entailing at least 3 contiguous voxels to identify a calcified lesion, which results in a minimum lesion area of 1.02 mm2. Splitting the coronary arterial tree into 17 segments, CCS was measured traditionally by the Agatston method  in MDCT images, and the visual method was applied as estimation method with ECG-triggered chest CTs.
ECG-triggered chest CT and Agatston method for CCS estimation
MDCT was made on all the population studied, and an Aquilion 64TM tomograph (Toshiba Medical Systems, Japan) was used for the acquisition of the images. Further, a prospective protocol was used, triggered by ECG, and an initial topogram to determine the limits of the cardiac volume to be acquired. Image reconstruction was made on 75% of the R-R interval; no iodinated contrast was administered, and the parameters established for the scanner were the following: collimation of 4 x 3 mm, tube voltage 120 Kv, tube current 300 mA, gantry rotation 250 ms, and visual field of 210 mm. The images were obtained at 3-mm intervals, starting 1 cm above the carina of the trachea, and progressing caudally until all coronary arteries are included. The effective dose of radiation with this technique was approximately 1 mSv per patient. The score for the calcified lesions was made in VitreaTM workstations, and the traditional Agatston method was used  for its evaluation, multiplying the calcification area by a factor derived from the maximum Hounsfield unit (HU) within the area (1 for lesions with maximum density between 130.199 HU; 2 for lesions between 200-299 HU; 3 for lesions between 300-399 HU; and 4 for lesions >400 HU). The Agatston score was divided into four groups (0, 1-100, 101-400 and >400).
ECG-triggered chest CT and visual method to estimate CCS
ECG-triggered chest CT was also made on the population studied, and the Aquilion LightningTM tomograph (Canon Medical Systems, Japan) was used to acquire images, which were obtained in helical mode, and the parameters established for the scanner were the following: tube voltage 100 Kv, tube current 430 mA, gantry rotation 500 ms, and visual field of 450 mm. The images were reconstructed as contiguous with a slice thickness of 3 mm to measure CCS; and to perform the visual method, one (1) point was assigned to each coronary artery segment that presented parietal calcification (17-segment division of all the arterial tree). The total visual estimation of CCS (TVCCS) for each individual was subjected to the sum of all the scores obtained in the four main arteries. TVCCS was categorized in the following manner: absent (0), mild (1-3), moderate (4-6) and severe (≥7), so as to match Agatston scores of 0, 1-100, 101-400 and ≥401, respectively (Figure 1).
Figure 1. A and C: MDCT, Agatston method. B and D: ECG-triggered chest CT, visual method.
All values were reported as mean and standard deviation ±(SD) and the concordance between the categorical variables of the visual method and the Agatston method, by a score of four scales, made through the Spearman’s rank correlation coefficient. Moreover, the variables between both methods were analyzed through the Chi-square test. Data were analyzed with computer programs devoted to this aim. A p value <0.05 was considered statistically significant.
Fifty patients were enrolled during the May-June 2017 term, with ages ranging from 35 to 87 years (average of 61 years). From the total population, 21 (42%) patients were males and 29 (58%) females, 23 (46%) hypertensive, 21 (42%) smokers, 17 (35%) dyslipidemic, and 4 (8%) diabetic, and 20 (40%) patients presented at least two risk factors for atherosclerotic CAD. The different CCS categories were examined according to the clinical limits usually used (0, 1-100, 101-400 and ≥401).
In Table 1, the results of CCS are presented according to the method used and category assigned. In the tests with ECG-gating, it was observed that 23 (46%) patients presented CCS of zero through the Agatston method; 15 (30%) had scores between 1-100; 7 (14%) 101-400; and 5 (10%) presented scores >400. In turn, non-ECG triggered CCS (visual method) revealed 24 (48%) patients with TVCCS of zero (equivalent to 0 by the Agatston method); 16 (32%) patients between 1-3 (equivalent to Agatston 1-100); 4 (8%) with TVCCS 4-6 (Agatston 101-400), and 6 (12%) patients with score ≥7 (Agatston >400). This means that 47 (94%) from the 50 cases evaluated presented agreement between both results, and in the 3 mismatching cases, the non-triggered tests showed a mild overestimation (Figure 2).
Table 1. CCS Agatston - Visual
Figure 2. Number of patients that presented a correlation between both methods (visual - Agatston).
Variability between triggered and non-triggered images increased with the highest CCS scores using any score method. Using Spearman’s rank correlation coefficient (Figure 3), it was demonstrated that the correspondence observed between the variables Agatston method and visual method, presented a statistically significant association; a value estimated of correlation r=0.97 (p<0.01). In turn, when using the Chi-square hypothesis test, a positive correlation between both methods used were also observed, with a proportion observed of 0.92 (p<0.01) (Table 2).
Figure 3. XY graph of the Spearman’s correlation test.
The value estimated for Spearman’s correlation is r=0.97 (p<0.01), significant association
between both variables (visual method and Agatston method).
Table 2. Chi-square test
Value under hypothesis
Visual –Agatston correlation
Rejection null hypothesis
This study showed an excellent correlation between CCS measurement in ECG-triggered chest CT by visual method, and the Agatston method obtained from MDCT. These findings will enable that in all ECG-triggered chest CT scans an objective evaluation will be possible for CAC, and in turn will provide a more accurate assessment about atherosclerotic load in the individuals evaluated through this type of imaging practice, in a similar manner to the one made by MDCT, thus allowing a cardiovascular risk stratification with no additional costs and no increase in exposition to ionizing radiation in the evaluated patients. Learning this visual score technique is very simple, allowing medical professionals, particularly imaging specialists and cardiologists that interpret this type of images, to have a simple and functional method to execute the principle of “accuracy medicine”; with its methodological simplicity being its great virtue, besides not requiring a dedicated workstation.
Although for a long time it was speculated that the capacity to quantify CAC accurately through non-triggered CAT could be impaired by cardiac motion, several studies started to show the concordance between both methods, and thus rule out such possibility. One of the first studies that investigated this topic was published in year 2008 by Wu et al . In this paper, the authors set out to determine the correspondence between CCS by ECG-triggered chest CT and MDCT in patients studied due to suspicion of lung cancer (483 patients, average age 62±13 years). All patients underwent both tomography scans, and two blind observers estimated CCS; in both cases through the area density method or the Agatston method . The correlation observed between both techniques and the four main score ranges (0, 1-100, 101-400, >400) was high (K=0.89 for the two observers).
Subsequently, in year 2010, a simpler method based on visual estimation, was proposed in the work by Shemesh et al . This investigation had a study population of 631 asymptomatic patients, and not just an excellent correlation was validated between this type of score with the classical Agatston score (r=0.84, p<0.0001), but also its significant predictive value in terms of cardiovascular death in individuals with visual CCS greater than 4 points (OR=4.7; 95% CI 3.3-6.8, p<0.0001). Another significant study on this issue was the one by Budoff et al , where the correlation between CAC scores between triggered and non-triggered MDCT were evaluated, in fifty patients enrolled in the Genetic Epidemiology of COPD (COPDGene) study. All patients underwent ECG triggered and non-triggered MDCT of 64-detector rows. Just as in our study, CAC was defined as vascular tissue with more than 130 HU in a minimum area of 1 mm2, and a slice thickness for both methods of 2.5 mm. The way to estimate CAC in non-triggered MDTC was the same as the Agatston method. The correlation between the mean values of CAC through both MDCT was excellent due to the intraclass correlation coefficient (r=0.96, ICC=0.96, p<0.0001), just as through the Bland-Altman graphs (mean difference 354, 95% CI, 168.538). The authors concluded in their work, that non-triggered MDCT is reliable to detect and quantify CAC.
Another significant aspect linked to the visual method sensitivity, is tomographic slice thickness. About this, Huang et al , studied the impact of the visual method (based on the length of the calcified lesions and their sum) and slice thickness (3 and 5 mm) in 401 individuals that underwent MDCT for the detection of lung cancer, and compared them with the results obtained in ECG-triggered MDCT and performance of the Agatston method. CCS by visual method and slice thickness of 3 mm was more concordant with the Agatston score than with 5 mm (k=0.813 vs k=0.685, respectively), just as the sensibility to detect CAC (83.6% vs 74.0%, respectively). Although thinner slices could be assumed to detect more CAC, this hypothesis could not be corroborated by Kim et al , on whose work more accuracy was determined (90%) to detect CAC in non-triggered MDCT with a slice thickness of 2.5 mm (72% and 84% for slices of 1 and 5 mm respectively); that is to say that thinner “slices” do not guarantee better results. Further, using a slice thickness of 3 mm would present less “partial volume” effect, which is not only advantageous for the estimation of CCS in non-ECG MDCT; but also for the diagnosis and follow-up of pulmonary nodules, as according to the recommendation by the Fleischner Society  from 2017, a minimum size of 6 mm is the criterion used to follow these cases over time.
Different authors  claim that a potential role of this type of test is to function as “doorman” for more advanced tests. The proposal would be that individuals with ECG-triggered chest CT and CCS by visual method of zero would not require a more dedicated CT; and in turn, those with mild values in the visual score (1-3) should be considered to undergo MDCT, and thus a proper estimation of the Agatston score; and finally, those with moderate (4-6) to severely high values (≥7) by visual method, should be evaluated by coronary angiography by MDCT or myocardial perfusion imaging under strain.
This study hast the merit of extending the subjective assessment of CAC, to a more quantitative measurement, much more accurate and comparable to the “gold standard” method than the mere dichotomous observation of the presence or not of coronary parietal calcium in ECG-triggered chest CT, giving a more objective tool to quantify CAC for the professional reporting or observing such images, and thus improving the risk evaluation of coronary events in the patient studied. Its second great merit lies in the potential to evaluate cardiovascular diseases without additional radiation or costs or complication, in any individual evaluated due to other extra-cardiac conditions; but all in all, a population with a potential risk of atheromatous CAD, enabling the detection of a “vulnerable” subject more precisely.
Limitations: Among them, the following stand out: First, most imaging specialists and cardiologists are not familiar with the visual evaluation of CAC in ECG-triggered chest CT; although this limitation was also found in other studies [13,18]; i.e. CAC and the visual method, after a short training on this methodology, an excellent correlation was achieved with the Agatston method, which entails a rapid learning of this new estimation method in professionals not familiar with this method.
Second, although there is still no standard method for the visual evaluation of CAC in ECG-triggered chest CT, the criteria of score proposed could be used as a reference for any professional interested in quantifying CAC in this type of CTs.
Third, densely calcified lesions in a single vessel or a single coronary segment, could underestimate the total value of CCS through the visual method; however, in this study, there were no mismatches between both methods in these individuals.
Fourth, the small size of the studied population, although the number of cases evaluated was enough to find a very good correlation between both methods.
This study shows the very good correlation existing between the method based on the visual observation of the images acquired in ECG-triggered chest CT and the traditional Agatston method.
These findings could be the basis to consider a widespread use of this methodology in the accurate evaluation of CAC in individuals investigated by suspicion of lung cancer or other chest pathologies, enriching the information provided about the patient studied.
The visual method described in this study, to quantify CAC, non-synchronized, ECG-triggered chest CT, could be applied in any patient with risk factors for atheromatous CAD.
Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990; 15: 827-32.
Hutt A, Duhamel A, Deken V, et al. Coronary calcium screening with dual-source CT: reliability of ungated, high-pitch chest CT in comparison with dedicated calcium-scoring CT. Eur Radiol. 2016; 26 (6): 1521-28.
Leening MJ, Elias-Smale SE, Kavousi M, et al. Coronary calcification and the risk of heart failure in the elderly: the Rotterdam Study. JACC Cardiovasc Imaging. 2012; 5 (9): 874-80.
Hermann DM, Gronewold J, Lehmann N, et al. Nixdorf Recall Study Investigative Group. Coronary artery calcification is an independent stroke predictor in the general population. Stroke. 2013; 44 (4): 1008-13.
Brenner DJ, Hall EJ: Computed tomography: an increasing source of radiation exposure. N Engl J Med. 2007; 357: 2277-84.
Williams Sr KA, Kim JT, Holohan KM. Frequency of unrecognized, unreported, or underreported coronary artery and cardiovascular calcification on noncardiac chest CT. J Cardiovasc Comput Tomogr. 2013; 7: 167-72.
Uretsky S, Chokshi N, Kobrinski T, et al. The interplay of physician awareness and reporting of incidentally found coronary artery calcium on the clinical management of patients who underwent noncontrast chest computed tomography. Am J Cardiol. 2015; 115: 1513-17.
Greenland P, Bonow RO, Brundage BH, et al. Coronary artery calcium scoring:ACCF/AHA2007 clinical expert consensus document on coronary artery calcium scoring by computed tomography in global cardiovascular risk assessment and in evaluation of patients with chest pain. J Am Coll Cardiol. 2007; 49: 378-402.
Greenland P, Alpert JS, Beller GA, et al. 2010 ACCF/AHA guideline for assessment of cardiovascular risk in asymptomatic adults: Executive Summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol. 2010; 56 (25): e50-103.
Kim SM, Chung MJ, Lee KS, et al. Coronary Calcium Screening Using Low-Dose Lung Cancer Screening: Effectiveness of MDCT with Retrospective Reconstruction. AJR Am J Roentgenol 2008; 190: 917-22.
Wu MT, Yang P, Huang YL, et al. Coronary arterial calcification on low-dose ungated MDCT for lung cáncer screening: concordance study with dedicated cardiac CT. AJR Am J Roentgenol 2008; 190: 923-28.
Shemesh J, Henschke CI, Shaham D, et al. Ordinal scoring of coronary artery calcifications on low-dose CT scans of the chest is predictive of death from cardiovascular disease. Radiology 2010; 257: 541-48.
Budoff MJ, Nasir K, Kinney GL, et al. Coronary artery and thoracic calcium on non-contrast thoracic CT scans: comparison of ungated and gated examinations in patients from the COPD Gene cohort. J Cardiovasc Comput Tomogr. 2011; 5: 113-18.
Huang YL, Wu FZ, Wang YC, et al. Reliable categorisation of visual scoring of coronary artery calcification on low-dose CT for lung cancer screening: validation with the standard Agatston score. Eur Radiol. 2013; 23 (5): 1226-33.
Mac Mahon H, Naidich DP, Goo JM, et al. Guidelines for management of incidental pulmonary nodules detected on CT images: from the Fleischner Society 2017. Radiology 2017; 284 (1): 228-43.
Rumberger JA. Using noncontrast cardiac CT and coronary artery calcification measurements for cardiovascular risk assessment and management in asymptomatic adults. Vasc Health Risk Manag 2010; 6: 579-91.
Einstein AJ, Johnson LL, Bokhari S, et al. Agreement of visual estimation of coronary artery calcium from low-dose CT attenuation correction scans in hybrid PET/CT and SPECT/CT with standard Agatston score. J Am Coll Cardiol 2010; 56: 1914-21.