Setting: The current study was conducted at Prince Sultan Cardiac Center (PSCC). The PSCC is 200-bed specialized cardiac center located in Riyadh that provides a major portion of the diagnostic and therapeutic cardiac services in Saudi Arabia. The PSCC has several departments including adult and pediatric cardiology, adult and pediatric cardiac surgery, cardiac anaesthesia, and advanced imaging. The current study was done at the advanced imaging unit under adult cardiology.
Design: It was a retrospective cohort study conducted between July 2007 and December 2017. The study design obtained all required ethical approvals from the ethical committee of PSMMC.
Population: The study targeted adult patients (age >18 years) referred to (64 multidetector spiral) computed tomographic (CT) for standard indications. Those with pre-existing CAD were excluded from the study. Pre-existing CAD was defined as myocardial infarction, angioplasty with stent placement and coronary artery bypass grafting. Additionally, CT done for TAVI, pericardial assessment and low-quality CT with artifacts were excluded from the study. Finally, those who were missing blood lipid (N=441) testing results were excluded, leaving 2421 for analysis.
CT scanning protocol: Patients were scanned during a single breath-hold using a 64 (multidetector spiral) CT scanner (Philips Brilliance). A retrospective gating protocol with thickness of 0.5 to 2.5, FOV 220, and the average radiation dose is 6-9 mSv. The scanning protocol was designed to minimize radiation dose based on BMI. Indications of coronary CT included chest pain in patients with intermediate risk of CAD, impaired left ventricular function in asymptomatic patient, before non-coronary cardiac surgery in patients with intermediate risk of CAD, to rule out coronary anomaly, and in case of arrhythmia with atypical chest pain.
Definitions: Plaque diagnosis was based on the results of CT scanning and post-test CT angiography. Soft and calcified plaques were confirmed when these is coronary stenosis (of any degree) on CT angiography with or without coronary calcification using CT scanning, respectively. Blood lipids results at the time of the computed tomography were used. The limits defining control of blood lipids (in mmol/L) were <5.17 for total cholesterol, ≥1.0 for males and ≥1.3 in females for HDL cholesterol, and <1.7 for triglycerides. The limit defining control of LDL cholesterol (in mmol/L) was <3 in mild SCORE risk, <2.6 in moderate SCORE risk, <1.8 in high SCORE risk, and <1.4 in very high SCORE risk.
Risk stratification: Stratification of the patient risk for CAD was done using the number of risk factors and the systematic coronary risk evaluation (SCORE) of the European Society of Cardiology (ESC). Risk factors for CAD were defined as history of hypertension, diabetes, dyslipidemia, smoking, family history of premature CAD (before the age of 65 years), and obesity (BMI>30). The SCORE estimating the 10-year risk of developing fatal cardiovascular disease in populations with high cardiovascular disease risk was calculated according to standard methodology [17].
Data collection tool: Study data collection sheet was initiated for patients who underwent coronary CT and meeting the study eligibility criteria. Clinical information including medical history, traditional risk factors, and cardiac comorbidity were then abstracted from the electronic patient chart system.
Statistical analysis: Demographic characteristics, risk stratification, and blood lipids were compared by the plaque and gender groups. Chi-square or Fisher exact tests (as appropriate) were used to detect differences in categorical variables. Analysis of variance (ANOVA) or Kruskal–Wallis test (as appropriate) were used to detect plaque differences in continuous variables while t-test or Mann–Whitney test (as appropriate) were used to detect gender differences in continuous variables. In case of plaque groups, pairwise differences were calculated using Bonferroni adjustment method for multiple comparisons. Generalized linear regression models was used to detect differences in log-transformed blood lipid levels between plaque and gender groups after adjusting for other CAD risk factors. Log data were transformed back to normal before presenting in Table 4. All P-values were two-tailed. P-value <0.05 was considered as significant. SPSS software (release 25.0, Armonk, NY: IBM Corp) was used for all statistical analyses.