Between March 2009 and March 2011, 586 consecutive Japanese patients underwent 64-slice cardiac CT (both CAC scoring and CTCA) for suspected coronary artery disease (CAD) at our hospital. We excluded patients with known CAD or previous myocardial infarction, those with a history of percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) surgery and those with unacceptable CTCA image quality. Among the remaining 353 patients, we enrolled 111 (31%) with zero CACS. This study was approved by our institutional review board, and all patients provided written informed consent to participate.

The patients were evaluated by cardiac CT scanning using a 64-slice MDCT (Aquillion 64, Toshiba Medical Systems, Tochigi, Japan). Before scanning, blood pressure and heart rate were measured, sublingual nitroglycerin was administered to all of the patients, and those with an initial heart rate above 65 beats per minute were administered with beta-blockers in the absence of contraindications. Thereafter, unenhanced electrocardiographically gated CT images were acquired to calculate CACS before CTCA, which were prospectively triggered at 75% of the RR interval and obtained under the following parameters: collimation, 4 × 3.0 mm; gantry rotation time, 0.25 sec; tube voltage, 120 kV; and tube current, 300 mA.

The CTCA scans were acquired as follows. Patients received an intravenous bolus of 0.9 mL/kg (body weight) of contrast media (Iopamiron 370 mg I/mL; Bayer HealthCare, Osaka, Japan) at a rate of 2.5 - 4.5 mL/sec, followed by 30 mL of saline at the same rate using a double-head injector (Dual shot™, Nemoto Kyorindou, Tokyo, Japan). A region of interest was placed within the ascending aorta, and the scan was started when the CT density reached 100 Hounsfield Units (HU) above the baseline CT density. Images were acquired under the following parameters: collimation, 64 × 0.5 mm; gantry rotation time, 0.35, 0.375 or 0.4 sec; tube voltage, 120 kV; tube current, 400 mA; beam pitch, 0.175 to 0.2. Raw datasets of the scans were reconstructed using retrospective electrocardiogram-gated halfscan or segmental reconstruction algorithms. The optimal cardiac phase was individually determined to minimize motion artifacts.

The reconstructed data were transferred to a computer workstation (Volume Analyzer SYNAPSE VINCENT, FUJIFILM, Tokyo, Japan) for CAC scoring and NCP evaluation.

We calculated CACS using the Agatston scoring method [21] and patients with zero CACS were enrolled in this study. All CTCA images were interpreted by two experienced observers who were blinded to the clinical findings of the patients such as their characteristics, cardiovascular risk factors, clinical presentations and pretest probabilities. If the CTCA findings of the observers differed, consensus was reached and the final findings were determined. Coronary arteries were divided into 15 segments as proposed for the American Heart Association model [22], and we evaluated all segments with a diameter of > 2 mm. Using curved multiplanar reformation (c-MPR) and cross-sectional (CS) images, we detected NCPs as structures of > 1 mm³ that could be assigned to the coronary artery walls with CT values below the contrast-enhanced coronary lumen but above the surrounding connective tissue, as previously described [23,24]. We then assessed the number of segments with NCP for each patient. Each NCP was characterized for remodeling index (RI), minimum CT value and the presence of spotty calcification (SC) and significant stenosis. We calculated RI from CS images as lesion vessel area/reference vessel area defined as the average of apparently normal proximal and distal vessel areas to the lesion. Positive remodeling (PR) was defined as RI > 1.1. At least 5 regions of interest were placed within each NCP, and the lowest value of those was defined as the minimum CT value. We defined NCP with a minimum CT value of < 0 HU as low CT value plaque (LP). We assessed the presence of NCPs with both PR and LP which might be associated with the subsequent development of ACS. We defined SC as being < 3 mm in size on c-MPR images and occupying only one side on CS images as described [7]. Stenosis was deemed significant when luminal diameter narrowing was ≥ 50%.

We assessed whether patients had the following cardiovascular risk factors based on their medical records: diabetes mellitus (DM), hypertension, dyslipidemia and current smoking. DM was defined as fasting glucose ≥ 126 mg/dL and hemoglobin A1c ≥ 6.5% or the use of antidiabetic treatment. Hypertension was defined as systemic blood pressure ≥ 140/90 mmHg or the use of antihypertensive treatment. Dyslipidemia was defined as total cholesterol ≥ 220 mg/dL, low-density lipoprotein cholesterol ≥ 140 mg/dL, fasting triglycerides ≥ 150 mg/dL, high-density lipoprotein cholesterol ≤ 40 mg/dL or the use of lipid-lowering treatment. Medicines administered at the time of CT imaging were also recorded. The pretest likelihood of CAD was determined based on Morise scores as low (< 9), intermediate (9 to15) or high (> 15) [25].

Continuous variables are expressed as means ± standard deviation (SD). Categorical variables are presented as absolute numbers and relative frequencies (%). The Chisquare test was used in univariate analyses for categorycal variables and Student’s unpaired t test was applied for continuous variables. Predictors of the presence of NCP, NCP with both PR and LP and ≥ 2 NCPs were assessed using multiple logistic regression analyses, and independent predictors of the number of segments with NCP were determined using multiple linear regression analysis. The diagnostic performance of the presence of NCP and ≥ 2 NCPs was evaluated using receiver operating characteristic (ROC) curve analyses. Cut off values were also calculated. All tests of significance were two-tailed and values of P < 0.05 were considered statistically significant. All statistical analyses were performed using Ekuseru-Toukei 2010 (Social Survey Research Information Co. Ltd., Tokyo, Japan).