Risk stratification in patients with acute coronary syndromes (ACS) is achieved today by clinical models, "blind" to the prognostic support of imaging methods. To assess the value of simple at rest cardiac chest sonography in predicting the intra- and extrahospital risk of death or myocardial infarction, we enrolled 470 consecutive in-patients (312 men, age 71 ± 12 years) who had been admitted for ACS. On admission, all had received a clinical score using the Global Registry in Acute Coronary Events and Thrombolysis in Myocardial Infarction systems and, within 1 to 12 hours, a comprehensive cardiac-chest ultrasound scan. Each of the 16 echocardiographic parameters evaluating left and right, systolic and diastolic, ventricular function and structure, was scored from 0 (normal) to 3 (severely abnormal). The median follow-up was 5 months (interquartile range 1 to 10). Patients with hard events (n = 102) could be separated from patients without events (n = 368) using the Global Registry in Acute Coronary Events score, Thrombolysis in Myocardial Infarction score, and several echocardiographic parameters. On multivariate Cox analysis, ejection fraction (hazard ratio 1.45, 95% confidence interval 1.02 to 2.08, p = 0.040), tricuspid annular plane systolic excursion (hazard ratio 1.66, 95% confidence interval 1.13 to 2.45, p = 0.010) and ultrasound lung comets (hazard ratio 1.69, 95% confidence interval 1.25 to 2.27, p = 0.001) were independent predictors of cardiac events. The 3-variable echocardiographic score (from 0, normal to 9, severe abnormalities in ejection fraction, ultrasound lung comets, and tricuspid annular plane systolic excursion) effectively stratified patients and added value (hazard ratio 2.52, 95% confidence interval 1.89 to 3.37, p <0.0001) to the Global Registry in Acute Coronary Events score (hazard ratio 1.60, 95% confidence interval 1.07 to 2.39, p = 0.003). In conclusion, for patients with ACS, effective risk stratification can be achieved with cardiac and chest ultrasound imaging parameters, adding prognostic value to the clinical risk scores.
Read full abstract