Quantification of coronary artery calcium has prognostic value and is commonly used in asymptomatic patients. Routine clinical use of coronary artery calcium in other populations remains uncertain. We sought to understand the potential application of the Agatston score in patients with heart failure (HF). For this purpose, 3 populations were identified: (1) patients with an Agatston score equal to 0, (2) patients with high-risk coronary artery disease (CAD) defined as 3-vessel, left main, or 2-vessel disease involving the proximal left anterior descending coronary artery, and (3) patients with HF symptoms and left ventricular (LV) ejection fraction <50%. Excluding patients with HF or LV dysfunction, 738 patients (mean age 52 ± 10 years, 43% men) had an Agatston score equal to 0. Of these, 18 (2%) had obstructive CAD (diameter stenosis ≥50%), 8 (1%) had diameter stenoses ≥70%, and none had high-risk CAD. The 74 patients with high-risk CAD without LV dysfunction had high Agatston scores (mean 895 ± 734, median 716, range 50 to 3,210). In total 153 patients with a history of HF and abnormal ejection fraction were identified. All 13 patients with ischemic cardiomyopathy had Agatston scores >0, whereas 46 of 140 patients (30.1%) with nonischemic causes had an Agatston score equal to 0. An Agatston score equal to 0 identified nonischemic causes with a specificity of 100% (confidence interval 90 to 100) and positive predictive value of 100% (confidence interval 90 to 100). Agatston score equal to 0 had incremental value to pretest probability for CAD. In conclusion, an Agatston score equal to 0 confers a very low likelihood of obstructive CAD, appears to rule out high-risk CAD, and thus may be used to rule out ischemic cardiomyopathy in patients with HF.
An Agatston score equal to 0 portends an excellent prognosis, but the clinical utility of coronary artery calcium (CAC) in patients with heart failure (HF) has not been well described. We sought to understand the potential application of the Agatston score in patients with HF by (1) determining the prevalence of nonobstructive, obstructive, and high-risk coronary artery disease (CAD) in a population of patients with an Agatston score equal to 0, (2) determining the distribution of Agatston scores in patients with high-risk CAD, and (3) assessing the potential application of the Agatston score in patients with HF and left ventricular (LV) systolic dysfunction. By understanding the characteristics of these populations, we endeavored to determine a clinical application for CAC assessment in patients with HF.
Methods
From February 2006 to February 2009, 4,394 patients underwent cardiac computed tomography and were prospectively enrolled into a registry that was approved by the institutional ethics research board. Pretest probability for obstructive CAD was calculated for individual patients using age, gender, and symptoms. In total 1,728 patients who underwent Agatston scoring and computed tomographic coronary angiography were screened. Patients with a history of coronary revascularization were excluded.
Heart rate–controlling medications were administered targeting ≤65 beats/min. In the absence of contraindications, patients received nitroglycerin 0.8 mg sublingually. For calcium scoring, a noncontrast-enhanced, prospective, electrocardiogram-gated image acquisition (400 to 800 mA, 120 kV) was performed at the 70% phase and images were reconstructed with a 2.5-mm slice thickness.
A timing bolus method was used to determine transit time. Final computed tomographic coronary angiographic images were acquired using a triphasic intravenous contrast administration protocol. Retrospective electrocardiogram-gated datasets were acquired with the GE Volume Computed Tomography Scanner (General Electric, Milwaukee, Wisconsin) with 64- × 0.625-mm slice collimation and a gantry rotation of 350 ms (400 to 800 mA, 120 kV, 0.16 to 0.24 pitch). Using electrocardiogram-gated x-ray tube modulation, x-ray output was minimized during systole. For evaluation of coronary arteries, datasets were reconstructed at 0.625 mm with an increment of 0.4 mm using the cardiac phase(s) with the least cardiac motion. For evaluation of LV volumes and ejection fraction (EF), images were reconstructed using a slice thickness of 1.25 mm and increment of 0.625 mm for 10 phases (5% to 95%).
Noncontrast and contrast images were processed using the GE Advantage Volume Share Workstation (GE Healthcare, Milwaukee, Wisconsin) and interpreted by expert observers blinded to all clinical data. Noncontrast studies were used to calculate the Agatston score.
Contrast-enhanced images were used to assess for nonobstructive and obstructive CAD using axial images, multiplanar reformations, and maximal intensity projections. A 17-segment model of coronary arteries and 4-point grading score (normal; mild, <50%; moderate, 50% to 69%; severe, ≥70%) were used to grade coronary stenosis. High-risk CAD was defined as left main coronary artery diameter stenosis ≥50%, 3-vessel disease (≥70% diameter stenosis in left anterior descending, left circumflex, and right coronary arteries), or 2-vessel disease (diameter stenosis ≥70%) involving the proximal left anterior descending coronary artery.
LV end-diastolic and end-systolic volumes were measured by cardiac computed tomography and LVEF was then calculated. An EF <50% was considered abnormal.
Three patient populations were identified: (1) patients with an Agatston score equal to 0 (without history of HF and LV dysfunction), (2) patients with high-risk CAD (without history of HF or LV dysfunction), and (3) patients with a history of HF and EF <50%. A diagnosis of ischemic cardiomyopathy was made if LV dysfunction was attributed to CAD (confirmed by invasive coronary angiography) with corresponding wall motion abnormalities.
Statistical analyses were performed using SAS 9.2 (SAS Institute, Cary, North Carolina), and statistical significance was defined as a p value <0.05. Continuous variables were presented as means ± SDs, and categorical variables were presented as frequencies with percentages. To compare patient characteristics and imaging parameters, Wilcoxon rank-sum test was used to compare continuous variables and Fisher’s exact test was used for categorical variables.
Diagnostic test characteristics (sensitivity, specificity, negative predictive value, and positive predictive value) were reported with 95% confidence intervals (CIs). Logistic regression models were used to evaluate the diagnostic value of pretest likelihood of CAD and addition of an Agatston score equal to 0 for ischemic cardiomyopathy. The penalized likelihood approach of Firth was applied in logistic models to correct for separability. Receiver operator characteristic curves were constructed for pretest probability of CAD alone and for a model of pretest probability of CAD and an Agatston score equal to 0. Area under 2 receiver operator characteristic curves was compared to evaluate the discrimination ability of each model to predict ischemic cardiomyopathy. The increased discriminative value of an Agatston score equal to 0 was further examined by estimating integrated discrimination improvement. To assess model calibration, the Hosmer-Lemeshow chi-square statistic was computed for all models.
Results
Of the 1,728 patients, 738 (mean age 51.7 ± 10.3 years, 43% men) without a history of HF or LV dysfunction had an Agatston score equal to 0 ( Table 1 ). Of these, 226 patients (31%) had coronary atherosclerosis and 36 patients (5%) had visual evidence of calcific plaque that was below the threshold of quantification using the Agatston method. Of patients with an Agatston score equal to 0, 208 (28%) had nonobstructive CAD (<50% diameter stenosis), 18 (2%) had obstructive CAD (diameter stenosis ≥50%), and 8 (1%) had diameter stenoses ≥70%. No patient had high-risk CAD ( Table 2 ) as defined by left main coronary artery disease ≥50%, 2-vessel disease (≥70%) involving the proximal left anterior descending artery, or 3-vessel disease (≥70%).
| Variable | Agatston Score = 0 | High-Risk CAD | HF |
|---|---|---|---|
| (n = 738) | (n = 74) | (n = 153) | |
| Age (years) | 51.7 ± 10.3 | 62.1 ± 9.7 † | 58.8 ± 11.7 † |
| Men | 317 (43%) | 58 (78%) ‡ | 87 (57%) ‡ |
| Pretest probability for coronary artery disease | 19.3 ± 24.0 | 39.3 ± 32.0 † | 30.1 ± 33.3 † |
| Cardiac risk factors | |||
| Hypertension | 305 (41%) | 57 (77%) † | 92 (60%) † |
| Diabetes mellitus | 75 (10%) | 21 (28%) † | 31 (20%) † |
| Current smoker | 102 (14%) | 14 (19%) | 19 (12%) |
| Ex-smoker >1 year | 233 (32%) | 24 (32%) | 69 (45%) † |
| Dyslipidemia ⁎ | 339 (46%) | 52 (70%) † | 85 (56%) ‡ |
| Family history of coronary artery disease | 331 (45%) | 35 (47%) | 5 (3%) |
| Indications for study | |||
| Chest pain | 442 (60%) | 43 (58%) | 76 (50%) ‡ |
| Nonanginal chest pain | 281 (38%) | 16 (22%) ‡ | 39 (25%) ‡ |
| Atypical chest pain | 103 (14%) | 14 (19%) | 16 (10%) |
| Typical chest pain | 58 (8%) | 13 (18%) ‡ | 21 (14%) ‡ |
| Dyspnea | 120 (16%) | 11 (15%) | 48 (31%) † |
| Palpitations | 45 (6%) | 5 (7%) | 8 (5%) |
| Syncope | 15 (2%) | 0 (0%) | 3 (2%) |
| Rule out coronary artery disease in asymptomatic patient | 35 (5%) | 9 (12%) ‡ | 9 (6%) |
| Equivocal stress test result | 57 (8%) | 6 (8%) | 1 (1%) † |
⁎ Self-reported dyslipidemia, use of lipid-lowering medications, and/or based on fasting lipid profile.
† p <0.001 compared to Agatston score = 0.
| Agatston Score = 0 | High-Risk CAD | HF | |
|---|---|---|---|
| (n = 738) | (n = 74) | (n = 153) | |
| Left ventricular ejection fraction (%) | 65 ± 7 | 66 ± 8 ⁎ | 45 ± 13% |
| Obstructive coronary artery disease ≥50% | 18 (2%) | 74 (100%) ⁎ | 50 (33%) ⁎ |
| Obstructive coronary artery disease ≥70% | 8 (1%) | 69 (93%) ⁎ | 34 (22%) ⁎ |
| Left main coronary artery ≥50% | 0 | 29 (39%) | 3 (2%) |
| 2-Vessel disease (≥70%) involving proximal left anterior descending coronary artery | 0 | 31 (42%) | 12 (8%) |
| 3-Vessel disease (≥70%) | 0 | 43 (58%) | 8 (5%) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
