Which is the best Myocardial Work index for the prediction of coronary artery disease? A data meta‐analysis

Abstract

Background: Early diagnosis of Coronary Artery Disease (CAD) plays a key role to prevent adverse cardiac events such as myocardial infarction and Left Ventricular (LV) dysfunction. Myocardial Work (MW) indices derived from echocardiographic speckle tracking data in combination with non-invasive blood pressure recordings seems promising to predict CAD even in the absence of impairments of standard echocardiographic parameters. Our aim was to compare the diagnostic accuracy of MW indices to predict CAD and to assess intra-and inter-observer variability of MW through a meta-analysis. Methods: Electronic databases were searched for observational studies evaluating the MW indices diagnostic accuracy for predicting CAD and intra-and inter-observer variability of MW indices. Pooled sensitivity, specificity, and Summary Receiver Operating Characteristic (SROC) curves were assessed.Results: Five studies enrolling 501 patients met inclusion criteria. Global Constructive Work (GCW) had the best pooled sensitivity (89%) followed by GLS (84%), Global Work Index (GWI) (82%), Global Work Efficiency (GWE) (80%), and Global Wasted Work (GWW) (75%). GWE had the best pooled specificity (78%) followed by GWI (75%), GCW (70%), GLS (68%), and GWW (61%). GCW had the best accuracy according to SROC curves, with an area under the curve of 0.86 compared to 0.84 for GWI, 0.83 for GWE, 0.79 for GLS, and 0.74 for GWW. All MW indices had an excellent intra-and inter-observer variability.Conclusions: GCW is the best MW index proving best diagnostic accuracy in the prediction of CAD with an excellent reproducibility.

From ejection fraction, to myocardial strain, and myocardial work in echocardiography: Clinical impact and controversies

“…3,4,10 On the other hand, relatively newer echocardiographic parameters (e.g., GLS and MW) for assessing LV global systolic function do not seem to be greatly helpful in the real-world routine clinical practice in the individual case, being currently more applicable to strictly selected patients at the population-level, excluding subjects affected by a large spectrum of pathologies (e.g., arrhythmias, severe valvular heart disease, prosthetic valve/s, hypertrophic cardiomyopathy, heart transplantation, left bundle branch block, and peripheral artery disease), 1,2 and strongly limited by the possible suboptimal ultrasound window as well. 1,2,5,14 Moreover, the wide cut-offs spread and overlap of values between groups, even though with (despite not always) statistically significant difference of mean values among different groups of patients in the published literature, has to be considered as a major limitation for the routine clinical application of the method. [1][2][3]5,[13][14][15][16][17][18][19][20][21][22][23][32][33][34][35] Effectively, the identification of decisional cut-off appears not realistic, and distinguishing the normal versus pathologic state may be difficult (or even impossible) in the individual case, because of overlap in the function values.…”

“…Moreover, MW analysis cannot be completely considered load‐independent, being derived by myocardial strain measurements as well 2 . Notably, each MW derived index would possibly include all variabilities and errors deriving from the measurements of each parameter, to be used for the final calculation 1,2 . Even with reported (despite not always) statistically significant difference of mean values among different groups of patients, furthermore, wide standard deviations and overlap have been observed, leading to a great difficulty in applying a univocal physiological meaning to a single MW value in a single patient, and universally acceptable precise cut‐offs of normality 32–35 .…”

“…Particularly in acute cardiovascular setting, including both acute HF and acute coronary syndromes, elevated pre and/or after‐load conditions are frequently present, as well as easily transient 4 . Theoretically, MW has the advantage to incorporate afterload information, through combining strain method with dynamic noninvasive LV pressure 1,2 (Figures 1, 2), in expenses of all the above‐mentioned intrinsic limitations. To date, MW analysis is considered useful in many clinical settings, including HF with reduced/preserved EF, coronary artery disease, valvular heart disease (e.g., aortic valvulopathy), diabetic and hypertensive disease, dilated cardiomyopathy, nonobstructive hypertrophic cardiomyopathy, amyloidosis with cardiac involvement, athlete's heart, and cardiac resynchronization 2,27–29 …”

“…Limitations : Accuracy 0.74 1 ; sensitivity 75% 1 ; specificity 61% 1 ; GLS‐related limits (including image quality); AP measurement‐related limits; single available software 2 …”

“…Notably, the precise pathophysiological meanings of these parameters can hardly be explained in different clinical contexts and situations. However, GCW currently seems to be the MW index of choice, mostly for the prediction of coronary artery disease 1 …”