The addition of rest left anterior descending coronary artery velocity to a clinical model significantly improved the prediction of obstructive CAD (AUC 0.875 vs 0.752; P<0.001).
Cohort (n=250)
Does the addition of rest LAD velocity by echocardiography to a clinical model improve the prediction of obstructive CAD in asymptomatic middle-aged subjects?
The addition of rest LAD velocity measured by Doppler echocardiography to standard clinical models significantly improves the identification of asymptomatic subjects with obstructive CAD.
Effect estimate: AUC 0.875 (95% CI 0.832 - 0.917)
p-value: p=<0.001
Abstract Background To date there is no published study investigating whether rest echocardiography, including left anterior descending coronary artery velocity Doppler sampling (LAD) and left ventricle global strain (GLS) in addition to standard measurements, can help predict which asymptomatic middle-aged subjects are affected by anatomically obstructive (stenosis 50%) coronary artery disease (CAD). In symptomatic subjects clinical risk factors and laboratory measurements are useful in the prediction of anatomically significant CAD. To date no single rest echocardiography variable has demonstrated additive value on top of clinical parameters in the prediction of CAD, assessed with CT coronary angiography (CTCA), in apparently healthy asymptomatic subjects. Methods Our screening program, named "pensiamoci-prima" ("think-earlier") was publicly advertised for voluntary healthy subjects with no known prior CAD, in the age range 45-69 living in the Venice area (Veneto, Italy) ; it prospectively enrolled more than 1000 subjects during 2 years, collecting comprehensive pre-determined clinical data, blood laboratory parameters, standard and advanced echocardiography variables, and also carotid intima-media thickness and the presence of carotid plaque using ultrasound. Echocardiography assessment was particularly thorough, and included GLS measurement and LAD velocity Doppler sampling in up to 3 sites (proximal, middle and distal LAD), whenever feasible. Patients instrumentally reclassified as higher-risk for CAD, based on the presence of at least one of the following pre-specified criteria: presence of carotid plaque, GLS-19 or LAD rest velocity (at any sampling site) 50 cm/sec, were suggested CAD screening with CTCA. While the project also tracks events and procedures during follow-up, we here report the preliminary results of the project regarding the diagnostic capability of clinical, laboratory and imaging variables to predict obstructive CAD, defined as at least one coronary stenosis 50%, in the 250 subjects who finally underwent CTCA. Results In the 250 subjects who underwent CTCA, 90 (36%) had at least one 50% stenosis. Among several clinical and imaging univariate predictors of CAD only few remained significant in the multivariate analysis (see Figure 1). Interestingly while a clinical model comprising demographics and risk factors showed AUC =0.752 (95% CI 0.691 - 0.813) , the simple addition of LAD velocity on top of the clinical model increased it significantly (p0.001) to AUC=0.875 (95% CI 0.832 - 0.917) (Figure 2). Sensitivity and specificity of the LAD-Clinical model were 67.8% and 90.6%, respectively. Conclusions This is the first study ever reporting that rest LAD velocity can be measured with high feasibility using Doppler echocardiography in a typical screening population of asymptomtic subjects, and may help select patients with highest probability of obstructive CAD who may benefit from non-invasive anatomic coronary imaging.Multivariate analysis to predict CAD Clinical model and +rest LAD velocity
Gaibazzi et al. (Sat,) conducted a cohort in Obstructive Coronary Artery Disease (n=250). Rest echocardiography with LAD velocity Doppler sampling vs. Clinical model (demographics and risk factors) was evaluated on Prediction of obstructive CAD (at least one coronary stenosis >50%) (AUC 0.875, 95% CI 0.832 - 0.917, p=<0.001). The addition of rest left anterior descending coronary artery velocity to a clinical model significantly improved the prediction of obstructive CAD (AUC 0.875 vs 0.752; P<0.001).