Incremental value of myocardial work analysis for detection of coronary artery disease during dipyridamole stress echocardiography

Abstract Funding Acknowledgements Type of funding sources: None. Background Dipyridamole stress-echocardiography (DipSE) relies on operator’s expertise and might have suboptimal diagnostic accuracy in identifying patients (pts) with single-vessel coronary artery disease (CAD). 2D strain, and, recently, myocardial work (MW) analysis, provide a more accurate evaluation of myocardial function, being able to detect subtle regional abnormalities. Purpose Aim of our study was to investigate whether MW analysis could have an additive role, when implemented to DipSE, in CAD detection. Methods We retrospectively enrolled 50 pts (20 M) who underwent DipSE followed, within one month, by coronary CT (CCT), in the absence of stress-induced WM abnormalities, or by invasive coronary angiography (ICA), when new WM abnormalities occurred. According to the presence of obstructive CAD (confirmed by ICA in any case), the population was subdivided into 2 groups: group 1 (with obstructive CAD: 34 pts) and group 2 (without obstructive CAD: 16 pts). After matching DipSE and CCT/ICA results, we further subdivided the population into three subgroups: 1 (true positives: 34 pts), 2A (false positives: 9 pts) and 2B (true negatives: 7 pts); no patients had a false negative DipSE result. Global longitudinal strain (GLS), work index (GWI), constructive work (GCW), waste work (GWW) and work efficiency (GWE) were calculated, at baseline, low and peak dose. Results In group 1, GLS increased at low and decreased at peak dose (p 0.015), while, in group 2, increased, mostly from low to peak dose (p< 0.001), Fig. 1. GWI and GCW, at low dose, decreased in group 1 (p 0.035 and p 0.009) whereas increased in group 2 (p 0.042 and p ns). GWW increased, at peak dose, in group 1 (p 0.003), whereas decreased in group 2 (p 0.047). GWE decreased, at peak dose, in group 1 (p 0.025), whereas increased in group 2 (p 0.001). When evaluating the subgroups (Fig. 1, right panels), we found that, in subgroup 2A, GLS increased at low and decreased at peak dose (p 0.015), similarly to subgroup 1 (at low dose p 0.011) and differently from subgroup 2B (from baseline to peak dose p 0.003). GWE, in subgroup 2A, increased at peak dose (p 0.023), differently from subgroup 1 (p 0.001) and similarly to subgroup 2B (at low dose: p 0.026; at peak dose: p 0.029). ROC analysis, for obstructive CAD, showed that GLS had a lower AUC (0.653) than GWW (0.735) and GWE (0.845), whereas GWI and GCW had, an AUC of 0.606 and 0.591, respectively (Fig. 2). A ≥ 0.5% GLS change (from low to peak dose) gained a sensitivity of 58%, a specificity of 67%, a positive predictive value (PPV) of 66% and a negative predictive value (NPV) of 64%. Conversely, a ≥ −1.5% GWE change (from low to peak dose) provided a sensitivity of 73%, a specificity of 84%, a PPV of 86% and NPV of 79%. Net reclassification improvement (NRI) for GWE was 0.049 (p 0.005). Conclusion MW analysis beyond GLS shows an additive role in the detection of CAD, thus improving diagnostic performance of DipSE.