Validation and Prognosis of Coronary Artery Calcium Scoring in Nontriggered Thoracic Computed Tomography

Abstract

Background— Coronary calcium score (CS), traditionally based on electrocardiography-triggered computed tomography (CT), predicts cardiovascular risk. Currently, nontriggered thoracic CT is extensively used, such as in lung cancer screening. The purpose of the study was to determine the correlation in CS between nontriggered and electrocardiography-triggered CT, and to evaluate the prognostic performance of the CS derived from nontriggered CT. Methods and Results— PubMed, Embase, and Web of Knowledge were searched until November 2012. Two reviewers independently screened 2120 records to identify studies reporting the CS in nontriggered CT and extracted information. Study quality was evaluated by standardized assessment tools. Cohen κ was extracted for agreement of CS categories between nontriggered and electrocardiography-triggered CT (validation). Hazard ratio (HR) was extracted for prognostic performance. Five studies about validation comprising 1316 individuals were included. Five studies about prognosis comprising 34 028 cardiac asymptomatic individuals, mainly from lung cancer screening trials, were included. All studies were of high quality. Meta-analysis could only be performed for validation studies because studies on prognostic performance were highly heterogeneous. Pooled Cohen κ for agreement between the 2 techniques was 0.89 (95% confidence interval, 0.83–0.95) for increasing CS categories. Increasing CS categories were associated with increasing risk of cardiovascular death or events. Nontriggered CT yielded false-negative CS in 8.8% of individuals and underestimated high CS in 19.1% of individuals. Conclusions— Our analysis shows the prognostic value and potential role of nontriggered assessment of coronary calcium, but it does not suggest that electrocardiography-triggered CT should be replaced by nontriggered examinations.