Summary
Background
The predictive value of CCTA to predict coronary artery disease is high in particular in the absence of coronary calcification. However, the consideration of both CCTA and the calcium score, in addition to the risk factors to determine the indication for coronary revascularization, has not been yet studied.
Materials and methods
This study included 2302 patients (mean age: 60 ± 9.8 years, 46% men), without known coronary artery disease (CAD), who underwent 320-row CCTA. Logistic regression, c-statistic and net reclassification improvement (NRI) were used to assess the role of coronary artery calcium score (CACS) in predicting revascularization after CCTA.
Results
The revascularization rates were 0.75% in patients with a CACS of 0, and there were no adverse events during the follow-up period. The revascularization rates were 3.3% in patients with a CACS of 1–99, 15.4% in patients with a CACS of 100–399, 25.6% in patients with a CACS of 400–999, and 42.4% in patients with a CACS ≥ 1000. The crude and adjusted odds ratios (95% confidence interval) for revascularization per CACS group category were 2.89 (2.53–2.3) and 2.71 (2.33–3.15), respectively; the area under the ROC curve (AUC) was 0.85 (0.83–0.88). The addition of CACS to conventional risk factors improved the accuracy of risk prediction model for revascularization (AUC 0.74 vs 0.63, P = 0.001), but it did not reclassify a substantial proportion of patients with positive CACS to risk categories (NRI = −0.023, P = 0.66).
Conclusions
The 320-row CCTA might rule out CAD in low- to intermediate-risk patients. However, its accuracy in identifying patients who require revascularization is limited. The CACS added to the conventional risk factors did not improve the identification of patients who require revascularization.
Résumé
Justification
La valeur prédictive du coronaroscanner pour éliminer une coronaropathie est élevée en particulier en l’absence de calcification coronaire. Cependant, la prise en compte simultanée des résultats du coronaroscanner, et de l’évaluation du score calcique en sus de la prise en compte des facteurs de risque pour déterminer l’indication à une revascularisation, n’a pas été étudiée à ce jour.
Méthodes
Cette étude a inclus 2302 patients, d’âge moyen 60 ± 9,8 ans, 46 % de sexe masculin, sans coronaropathie connue qui ont bénéficié d’une coronaroscanner. La régression logistique, le score C-statistic et l’amélioration nette de la reclassification (NRI) ont été utilisés pour évaluer le rôle du score calcique dans la prédiction de l’indication à une revascularisation au décours d’un coronaroscanner.
Résultats
Le taux de revascularisation était de 0,75 % chez les patients ayant un score calcique = 0 et il n’y avait pas d’évènement indésirable colligé pendant la période de suivi. Le taux de revascularisation était de 3,3 % chez les patients ayant un score calcique entre 0 et 99, 15,4 % lorsque ce score était entre 100 et 399, 25,6 % lorsque ce score était entre 400 et 999 et 42,4 % chez les patients ayant un score ≥ 1000. Les odds ratios brut et ajusté avec IC 95 % pour l’indication à la revascularisation par catégorie de score calcique étaient de 2,89 (IC 95 % −2,3–2,53) et de 2,71 (IC 95 % 2,33–3,15), respectivement. La surface sous la courbe ROC était à 0,85 (IC 95 % 0,83–0,88). Le fait de prendre en considération les facteurs de risque en sus du score calcique améliore la précision du modèle pour prédire l’indication à une revascularisation puisque la surface sur la courbe ROC passe de 0,63 à 0,74, p = 0,001. En revanche, il n’y a pas d’amélioration de la classification correcte des patients en utilisant le résultat du score calcique par catégorie, puisque le NRI est égal à −0,023, p = 0,66.
Conclusion
Le coronaroscanner peut éliminer une coronaropathie chez les patients à faible risque où à risque intermédiaire. Cependant, sa capacité à identifier les patients ayant une indication à une revascularisation est limitée. Le score calcique améliore la prédiction d’un geste de revascularisation, en sus de la prise en compte des facteurs de risque cardiovasculaire mais n’améliore pas l’identification des patients ayant une indication à revascularisation.
Background
320-detector row coronary computed tomography angiography (CCTA) has become a valuable imaging method for patients with suspected coronary artery disease (CAD), to avoid unnecessary invasive coronary angiography (ICA) . However, one of the major limitations of CCTA is the discrepancy between angiographic and functional stenosis . Coronary artery calcification is a marker of atherosclerosis that can be assessed using non-contrast CCTA. While the coronary artery calcium score (CACS) can assist the clinician in effectively ruling out angiographically significant CAD , the absence of coronary calcium does not preclude the presence of significant CAD with non-calcified plaques . Therefore, all patients with a CACS of 0 might benefit from further evaluation with contrast-enhanced CCTA . Previous studies showed that the CACS and CCTA had high sensitivity, but a much lower specificity for obstructive CAD in symptomatic patients with no prior diagnosis of CAD . Limited data exist on the assessment of ICA rates and revascularization in relation to CCTA results. Accordingly, we aimed to investigate the use of ICA and revascularization after CCTA in symptomatic patients at low to intermediate risk of CAD; and the relationship between the CACS with obstructive CAD and major adverse cardiac events (MACE) in patients with a positive CACS.
Methods
Study population
The present single-center study included consecutive symptomatic patients with a low to intermediate likelihood of CAD who underwent 320-detector row CCTA as a first-choice non-invasive test at the Silkeborg Diagnostic Centre, Denmark, from January 2011 to June 2013. Patients with a history of chest pain or other symptoms suggestive of CAD were included. The exclusion criteria were known CAD and CCTA scans with non-diagnostic image quality ( n = 108). Clinical characteristics, including age, sex, body mass index, cardiovascular risk factors (diabetes mellitus, use of antihypertensive and cholesterol-lowering drugs, smoking status, renal function), ICA and revascularization information were obtained from the Western Denmark Heart Registry, to which all patients were reported at the time of the CCTA. Follow-up data regarding MACE and all-cause deaths were obtained from the Danish National Patient Registry and patients’ health records. The study was approved by the Danish Data Protection Agency.
Computed tomography scan protocol and investigation reporting
All CACS and CCTA scans were performed using a 320-slice multislice computed tomography scanner (Aquilion One; Toshiba Medical Systems, Tokyo, Japan). Before imaging, patients with a heart rate > 65 beats/min were given metoprolol tartrate (Seloken ® 2.5–20 mg, single dose, intravenously; AstraZeneca) in the scanner room. Scanning for CACS was performed with a tube voltage of 120 kV, a tube current of 80 mA and a slice thickness of 0.5 mm. Patients were given a sublingual nitroglycerine spray (800 μg) 2 minutes before the CCTA scan, for coronary artery vasodilatation. A contrast bolus of 60–90 mL non-ionic contrast material (Iomeron ® 350 mg/mL; Bracco Imaging Scandinavia AB, Hisings Backa, Sweden) was administered at a flow rate of 5–6 mL/s, followed by a 40–50 mL saline chaser. Automated detection of peak enhancement in the aortic root was used to time the scan. A prospective electrocardiogram-triggered scan was performed at tube voltages between 100 and 135 kV, adapted to body mass index and thoracic anatomy, with an effective tube current of 100–580 mA, 0.5 mm slice thickness reconstruction and a gantry rotation time of 350 ms. All images were acquired during tidal inspiration, with patients holding their breath for approximately 10 seconds, and with simultaneous registration of the patient’s electrocardiogram. Before acquisition, breath-holding and heart rate tests were performed. The radiation dose was quantified with a dose-length product conversion factor of 0.014 mSv/(mGy × cm).
Images were initially reconstructed at 75%, and the best phase of the R–R interval using a soft-tissue kernel algorithm was optimized for cardiac imaging. In the case of motion artefacts and heart arrhythmia, a representative single slice was reconstructed throughout the cardiac cycle in steps of 10–20 ms, to determine the most optimal additional reconstruction phases. Subsequently, datasets were reconstructed and transferred to a remote Vitrea Workstation (Vitrea software, version 6.5.1, Vital images) for further diagnostic workup.
Image analysis
Independent observers with level 2 experience in CCTA performed the CCTA image analysis. The CACS was assessed with the VitreaWorkstation, using CACS analysis software. Coronary calcium was defined as an area of at least three “face-connected” voxels of peak density ≥ 130 HU within a coronary artery, corresponding to a minimum lesion area > 1 mm 2 , which was used as the reference value for the calcium scores. All lesions were added to calculate the Agatston calcium score. The total CACS was used in the analysis, and was determined by adding all lesions from each of the four main coronary arteries: left main coronary artery; left anterior descending coronary artery; left circumflex coronary artery; and right coronary artery. Patients with a high CACS (≥ 1000) did not undergo contrast CCTA, and were referred for ICA.
For all coronary artery segments, axial and multiplanar reformatted reconstruction images were created. Coronary anatomy was assessed in a standardized manner, by dividing the coronary artery tree into 17 segments based on the modified American Heart Association classification . A significant lesion was defined as a stenosis of ≥ 50% in the luminal diameter of the left main coronary artery or of ≥ 50–70% in the major epicardial coronary artery. The presence of significant lesions was determined based on visual estimation.
A three-point scoring system was used for image quality evaluation for each coronary artery. Scores defined as 1 had good image quality with no motion artefacts; scores defined as 2 had moderate image quality, acceptable for clinical diagnosis; and scores defined as 3 had poor image quality, with diagnosis being impossible because of severe artefacts. Only CCTA scans with scores of 1 or 2 were included in the study.
Outcomes
Referral for ICA and revascularization within 90 days after CCTA was considered to be an endpoint. The secondary endpoints were the incidences of MACE, defined as non-ST-segment elevation myocardial infarction, unstable angina requiring revascularization and cardiovascular death, and all-cause death during follow-up to a medium of 27 months (interquartile range 21–34 months).
Statistical analysis
Continuous variables are presented as means ± standard deviations; categorical variables are presented as percentages. The Kruskal–Wallis test was used to examine the mean differences within groups. Frequencies were compared using the χ 2 test for categorical variables. Total CACS values were classified into five categories: 0 (calcium absent); 1–99; 100–399; 400–999; and ≥ 1000. Spearman’s rho was used to assess the relationship between obstructive CAD on CCTA and revascularization. Logistic regression was used to identify the role of CACS in predicting revascularization within 90 days after CCTA. Adjustments were made for conventional risk factors: age, sex, family history of CAD, diabetes mellitus, current smoking and use of cholesterol-lowering and antihypertensive drugs. Multivariable regression models with and without CACS were composed to predict revascularization. Receiver operating characteristic (ROC) curves were used to assess the accuracy of the models for predicting revascularization. Significant differences in the areas under the ROC curves (AUCs) were compared using MedCalc statistical software. Net reclassification improvement (NRI) was based on the reclassification tables, and was calculated from the sum of the differences between the “upward” movement in categories for event subjects and the “downward” movement of non-event subjects . Cox regression was used to examine an association between CACS and MACE, and all-cause death. Ninety-five percent confidence intervals (CIs) were calculated for each comparison. A P -value < 0.05 was considered statistically significant. All tests were two-tailed. The Statistics Package for Social Sciences (SPSS) for Windows, version 17.0 (IBM, Armonk, NY, USA) was used for the analysis.
Results
Patient characteristics
The study population consisted of 2302 patients who underwent CCTA and matched the inclusion criteria. The mean age of the study cohort was 60 ± 9.8 years; 46% were men. During CCTA image acquisition, a mean heart rate of 56 ± 8 beats per minute was recorded. The estimated average radiation dose for the CCTA protocol was 3.5 ± 2.8 mSv (0.47 ± 0.02 mSv for CACS scans), using a conversion coefficient k of 0.014 for the chest; the mean contrast dose was 71 ± 12 mL. Of our 2302 study patients, 1064 patients had a CACS of 0, and 1238 patients had positive CACS values. The prevalence of a positive CACS was 64.5% (673/1042) for men, which was significantly higher ( P < 0.001) than that for women (44.8%; 565/1260). After CCTA, obstructive CAD was excluded in 1804 patients (78%). The remaining 498 patients (22%) with obstructive CAD on CCTA ( n = 393) and CACS ≥ 1000 ( n = 105) were referred for ICA, but only 187 (8.1%) underwent revascularization within 90 days after CCTA (a total of 137 percutaneous coronary intervention and 55 coronary artery bypass graft procedures). Nine patients underwent both percutaneous coronary intervention and coronary artery bypass graft surgery, and the remaining patients were recommended for conservative medical treatment.
Patient characteristics according to the CACS
Table 1 presents the baseline characteristics of the study population, distributed according to the CACS groups. The patients with higher CACS values were significantly older and were men. There was a significantly higher prevalence of diabetes mellitus, current smoking and use of antihypertensive and cholesterol-lowering treatments in patients with higher CACS values ( Table 1 ). We found no significant differences between these groups concerning family history of CAD, blood pressure, creatinine concentrations or body mass index.
| CACS groups | P | |||||
|---|---|---|---|---|---|---|
| 0 ( n = 1064) | 1–99 ( n = 541) | 100–399 ( n = 364) | 400–999 ( n = 215) | ≥ 1000 ( n = 118) | ||
| Age (years) | 54 ± 9.1 | 61 ± 9.5 | 64 ± 8.4 | 66 ± 7.3 | 67 ± 7.4 | < 0.0001 |
| Men | 35 | 51 | 52 | 64 | 61 | < 0.0001 |
| Family history | 48 | 46 | 56 | 50 | 47 | 0.051 |
| Diabetes | 7.1 | 8.6 | 15.5 | 12.6 | 19.5 | < 0.0001 |
| Current smoking | 17 | 18 | 23 | 22 | 29 | 0.008 |
| Systolic blood pressure (mmHg) | 135 ± 18 | 140 ± 19 | 140 ± 17 | 141 ± 17 | 144 ± 22 | 0.125 |
| Diastolic blood pressure (mmHg) | 82 ± 910 | 82 ± 9.5 | 82.5 ± 10 | 82.5 ± 10 | 82 ± 10 | 0.807 |
| Body mass index (kg/m 2 ) | 26.5 ± 4.6 | 26.7 ± 4.3 | 26.7 ± 4.3 | 26.9 ± 4.2 | 26.9 ± 4.8 | 0.84 |
| Creatinine concentration (μm/L) | 72 ± 14 | 75 ± 14 | 76 ± 16 | 78 ± 16 | 78 ± 18 | 0.404 |
| Ejection fraction | 59.5 ± 5.3 | 60.5 ± 5.4 | 59.8 ± 4.6 | 59.6 ± 5.6 | 58.5 ± 6.6 | 0.004 |
| Antihypertensive drug use | 30 | 42 | 48 | 56 | 60 | < 0.0001 |
| Cholesterol-lowering drug use | 25 | 38 | 42 | 49 | 58 | < 0.0001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
