The accuracy of multidetector computed tomographic (CT) angiography involving 64 detectors has not been well established.

Rationale:: Machine learning may be useful to characterize cardiovascular risk, predict outcomes, and identify biomarkers in population studies. Objective:: To test the ability of random survival forests, a machine learning technique, to predict 6 cardiovascular outcomes in comparison to standard cardiovascular risk scores. Methods and Results:: We included participants from the MESA (Multi-Ethnic Study of Atherosclerosis). Baseline measurements were used to predict cardiovascular outcomes over 12 years of follow-up. MESA was designed to study progression of subclinical disease to cardiovascular events where participants were initially free of cardiovascular disease. All 6814 participants from MESA, aged 45 to 84 years, from 4 ethnicities, and 6 centers across the United States were included. Seven-hundred thirty-five variables from imaging and noninvasive tests, questionnaires, and biomarker panels were obtained. We used the random survival forests technique to identify the top-20 predictors of each outcome. Imaging, electrocardiography, and serum biomarkers featured heavily on the top-20 lists as opposed to traditional cardiovascular risk factors. Age was the most important predictor for all-cause mortality. Fasting glucose levels and carotid ultrasonography measures were important predictors of stroke. Coronary Artery Calcium score was the most important predictor of coronary heart disease and all atherosclerotic cardiovascular disease combined outcomes. Left ventricular structure and function and cardiac troponin-T were among the top predictors for incident heart failure. Creatinine, age, and ankle-brachial index were among the top predictors of atrial fibrillation. TNF-α (tissue necrosis factor-α) and IL (interleukin)-2 soluble receptors and NT-proBNP (N-Terminal Pro-B-Type Natriuretic Peptide) levels were important across all outcomes. The random survival forests technique performed better than established risk scores with increased prediction accuracy (decreased Brier score by 10%–25%). Conclusions:: Machine learning in conjunction with deep phenotyping improves prediction accuracy in cardiovascular event prediction in an initially asymptomatic population. These methods may lead to greater insights on subclinical disease markers without apriori assumptions of causality. Clinical Trial Registration:: URL: http://www.clinicaltrials.gov . Unique identifier: NCT00005487.

Emerging evidence suggests that remnant cholesterol (RC) promotes atherosclerotic cardiovascular disease (ASCVD). We aimed to estimate RC-related risk beyond low-density lipoprotein cholesterol (LDL-C) and apolipoprotein B (apoB) in patients without known ASCVD. We pooled data from 17 532 ASCVD-free individuals from the Atherosclerosis Risk in Communities study (n = 9748), the Multi-Ethnic Study of Atherosclerosis (n = 3049), and the Coronary Artery Risk Development in Young Adults (n = 4735). RC was calculated as non-high-density lipoprotein cholesterol (non-HDL-C) minus calculated LDL-C. Adjusted Cox models were used to estimate the risk for incident ASCVD associated with log RC levels. We also performed discordance analyses examining relative ASCVD risk in RC vs. LDL-C discordant/concordant groups using difference in percentile units (>10 units) and clinically relevant LDL-C targets. The mean age of participants was 52.3 ± 17.9 years, 56.7% were women and 34% black. There were 2143 ASCVD events over the median follow-up of 18.7 years. After multivariable adjustment including LDL-C and apoB, log RC was associated with higher ASCVD risk [hazard ratio (HR) 1.65, 95% confidence interval (CI) 1.45–1.89]. Moreover, the discordant high RC/low LDL-C group, but not the low RC/high LDL-C group, was associated with increased ASCVD risk compared to the concordant group (HR 1.21, 95% CI 1.08–1.34). Similar results were shown when examining discordance across clinical cutpoints. In ASCVD-free individuals, elevated RC levels were associated with ASCVD independent of traditional risk factors, LDL-C, and apoB levels. The mechanisms of RC association with ASCVD, surprisingly beyond apoB, and the potential value of targeted RC-lowering in primary prevention need to be further investigated.

Background— Age-related alterations of left ventricular (LV) structure and function that may predispose to cardiovascular events are not well understood. Methods and Results— We used cardiac MRI to examine age-related differences in LV structure and function in 5004 participants without overt cardiovascular disease when enrolled in the Multi-Ethnic Study of Atherosclerosis; 1099 participants received additional strain analyses by MRI tagging. We also assessed the relation of age-associated remodeling with cardiovascular outcomes using Cox proportional hazard models adjusting for cardiovascular risk factors. Although LV mass decreased with age (−0.3 g per year), the mass-to-volume ratio markedly increased (+5 mg/mL per year, P <0.0001), driven by a substantial reduction in end-diastolic volume (−0.8 mL per year, P <0.0001). Age was also associated with a significant fall in stroke volume (−0.4 mL per year, P <0.0001), along with strain patterns reflecting systolic ( P <0.0001) as well as diastolic ( P <0.01) myocardial dysfunction—despite a modestly enhanced ejection fraction (+0.1% per year, P <0.0001). Increased mass-to-volume ratio conferred a significant risk for total cardiovascular events; this trend was strongest among younger (<65 years; hazard ratio, 3.69 [CI, 1.34 to 10.10]) versus older (≥65 years; hazard ratio, 1.68 [CI 0.77 to 3.68]) individuals with the highest compared to lowest mass-to-volume ratio quintile (P _interaction =0.013). Conclusions— Age is associated with a phenotype of LV remodeling marked by increased mass-to-volume ratio and accompanied by systolic as well as diastolic myocardial dysfunction that is not reflected by preserved ejection fraction. This pattern of ventricular remodeling confers significant cardiovascular risk, particularly when present earlier in life.

Aims: To evaluate the diagnostic power of integrating the results of computed tomography angiography (CTA) and CT myocardial perfusion (CTP) to identify coronary artery disease (CAD) defined as a flow limiting coronary artery stenosis causing a perfusion defect by single photon emission computed tomography (SPECT).Methods and results: We conducted a multicentre study to evaluate the accuracy of integrated CTA–CTP for the identification of patients with flow-limiting CAD defined by ≥50% stenosis by invasive coronary angiography (ICA) with a corresponding perfusion deficit on stress single photon emission computed tomography (SPECT/MPI). Sixteen centres enroled 381 patients who underwent combined CTA–CTP and SPECT/MPI prior to conventional coronary angiography. All four image modalities were analysed in blinded independent core laboratories. The prevalence of obstructive CAD defined by combined ICA–SPECT/MPI and ICA alone was 38 and 59%, respectively. The patient-based diagnostic accuracy defined by the area under the receiver operating characteristic curve (AUC) of integrated CTA–CTP for detecting or excluding flow-limiting CAD was 0.87 [95% confidence interval (CI): 0.84–0.91]. In patients without prior myocardial infarction, the AUC was 0.90 (95% CI: 0.87–0.94) and in patients without prior CAD the AUC for combined CTA–CTP was 0.93 (95% CI: 0.89–0.97). For the combination of a CTA stenosis ≥50% stenosis and a CTP perfusion deficit, the sensitivity, specificity, positive predictive, and negative predicative values (95% CI) were 80% (72–86), 74% (68–80), 65% (58–72), and 86% (80–90), respectively. For flow-limiting disease defined by ICA-SPECT/MPI, the accuracy of CTA was significantly increased by the addition of CTP at both the patient and vessel levels.Conclusions: The combination of CTA and perfusion correctly identifies patients with flow limiting CAD defined as ≥50 stenosis by ICA causing a perfusion defect by SPECT/MPI.