P1377 Predicting the presence of coronary artery disease by transesophageal echocardiography

Introduction The accuracy of ultrasound signs as predictors for the presence of coronary artery disease (CAD) has been evaluated extensively in the 1990s and 2000s. Imaging quality has improved tremendously over the last decades. Hypothesis High-end ultrasound systems allow for accurate prediction of presence or absence of CAD. Methods We retrospectively evaluated all patients who received transesophageal echocardiography examination (TEE) between 2007 and 2016 and who had coronary angiography within 24 months before or after the TEE. Results A total of 242 patients fulfilled the inclusion criteria, 60% were male. Mean age was 70 (SD ±13) years. In multivariate regression analysis, plaque in the aortic arch (OR 2.53, 95%CI 1.01-6.35, p = 0.048), plaque in at least one of the thoracic aortic segments (OR 2.44, 95%CI 1.21-4.9, p = 0.012), the presence of mitral annular calcification (MAC) (OR 1.87, 95%CI 1.03-3.39, p = 0.038), and any two of the three (aortic stenosis (AS), MAC, aortic plaque) in combination (OR 2.05, 95%CI 1.08-3.88, p = 0.028) were predictors of significant CAD. The isolated findings MAC (OR 1.91, 95%CI 1.00-3.65, p = 0.049) and AS (OR 2.45, 95%CI 1.22-4.92, p = 0.012) were significant predictors for the absence of normal coronary arteries. Conclusion With a negative predictive value of 80%, the absence of MAC, AS, and aortic plaque makes the presence of significant CAD unlikely. If CAD must be excluded and MAC or AS are present, there is high likelihood for at least mild CAD. Coronary angiography rather than coronary CT angiography should be applied in these patients.