The SMART-ORACLE Study

Patients with a prior cardiovascular event exhibit an elevated risk for subsequent cardiovascular events. The Second Manifestation of Arterial Disease Study (SMART) has recruited over 10,000 patients since 1996 with clinically manifest cardiovascular disease in a multidisciplinary single center study involving primary care physicians, cardiologists, neurologists, vascular surgeons, vascular medicine specialists and radiologists. Within this study a prediction model was developed to accurately estimate the risk for new cardiovascular events. It is now possible to acquire high-quality motion-free computed tomography (CT) images of the coronary and carotid arteries. Imaging biomarkers extracted from these images may further improve the prediction of future manifestations of arterial disease and personalize disease monitoring and medical care. The SMART-ORACLE study is a prospective, single center, observational cohort study aiming to include 1500 patients. The primary aim is to identify predictors of future cardiovascular events. Eligible patients will be selected via the recruitment of the original SMART study. Patients participating in the SMART-ORACLE study will undergo calcium scoring in multiple cardiovascular beds and contrast-enhanced CT-scans of the coronary and carotid arteries in addition to the regular SMART investigations. Patients with renal dysfunction will be either excluded from the study or will receive pre-hydration (depending on eGFR) to minimize the risk of contrast nephropathy. Follow-up with questionnaire-based assessment will take place every 26 weeks until death of participant or end of the SMART study, asking participants about possible new cardiovascular events. Endpoints will be adjudicated by a committee of three experts. The aim is to collect 170 future events. The main analysis will consist of Cox proportional hazard analysis. Imaging biomarkers will be added to the existing prediction model to assess their (independent) discriminatory capacity for future events. The c-statistic will be used to measure the discrimination of each model. Net reclassification improvement tables will be constructed to evaluate the added value of imaging markers in terms of reclassification. Based on a one-year-occurrence rate of subsequent cardiovascular events in the current SMART population database of 2.6%, about 6500 person years of follow-up are needed to obtain the adequate number of endpoints.