In the present study, we demonstrated that myocardial function is impaired at rest in most CAD patients. The primary findings of the study were: (i) Most global and regional MWs were of good diagnostic value for CAD (≥ 50% stenosis); (ii) GWI provided incremental value over all other echocardiographic parameters in predicting CAD, including GLS and LVEF. GWI might be a useful tool for risk stratification of patients with suspected CAD; (iii) Our data suggested that MWs in LAD and RCA segments had higher diagnostic value than MWs in LCX territory. The diagnostic power of regional MWs may be closely related to the location of severe stenosis (≥ 75% stenosis).
To our knowledge, few previous studies have evaluated the diagnostic value of MW in CAD[3, 8, 19]. Moreover, most of these reports have defined CAD patients as having ≥ 70% stenosis in one or more coronary arteries. Our study specifically evaluated MWs in patient subsets that included moderate CAD (stenosis ≥ 50%). And, the territorial correlation between MWs and angiography has rarely been assessed, because of lacking sufficient number of patients[8]. None of these researches have explored the role of regional MWs in predicting moderate, clinically unapparent CAD. In agreement with previous records[3, 8], GWI was demonstrated to be the most significant predictor in all MW indices for detecting moderate stenosis. When MWs were compared with GLS in this cohort, we found that the assessment of GWI (AUC = 0.760) was superior to the assessment of GLS (AUC = 0.690) and the diagnostic accuracy of GWI was better in terms of sensitivity and specificity. For the definition of CAD in our cohort is 50% stenosis, the diagnostic accuracy of GWI in our cohort shows relatively lower AUC compared with GWI (AUC = 0.786–0.806) obtained from two other cohorts whose classification standard is ≥ 70% stenosis[8, 19].
An optimal GWI cutoff value of 1855.5 mmHg % (sensitivity, 71.7%; specificity, 78.0%) has been defined in our cohort. In previous studies, the optimal cutoff GWI value to predict CAD (≥ 70% stenosis) were 1810 mmHg% (sensitivity, 92%; specificity, 51%) and 1790 mmHg% (sensitivity, 87.2%; specificity, 63.8%)[8, 19]. In addition, Boe et al.[3] found that regional MWI value of ༜1700 mmHg% gave the largest AUC at predicting patients with acute coronary artery occlusion. GWI values less than 1855.5 mmHg% may be commonly associated with moderate and severe coronary artery stenoses in patients with clinically suspected CAD. Cutoff points of MWs can be powerful tools for detecting CAD in patients with apparently normal ejection fraction and wall motion.
In the present study, the sensitivities of MWs (ranging from 60.4–71.7%) are slightly lower than the range reported in previous MW studies[3, 8]. Lower rates may be explained by the fact that the definition of significant coronary disease in these studies are different. For coronary stenosis ≥ 70% is related with more severe cardiac function impairment than stenosis ≥ 50%. Moreover, lower sensitivity may due to strict exclusion criteria. The enrolled patients were at relatively low risk of CAD. Nevertheless, the fact that GWI was a strong predictor of CAD even in a low-risk population implies that GWI might be a useful tool for risk stratification of patients with suspected CAD.
To our knowledge this represents the first investigation that specifically evaluates regional MWs in patient subsets that include moderate CAD. Previous studies have demonstrated a good correlation between the myocardial ischemia area and regional cardiac dysfunction[18, 20]. In the present study, most regional MWs were significantly impaired in segments supplied by stenotic coronary arteries compared with non-stenotic coronary arteries. So, it is possible to predict which coronary artery is stenotic. Regional analysis by 2D-STE has preliminarily diagnosed CAD patients with a good sensitivity. The specificity, however, was low which precludes its clinical use [21]. The MW measurements of our study showed relatively good specificities at identifying patients with coronary stenosis in LAD, LCX and RCA. This may partly due to more accurate analyses of MW than myocardial strain in evaluating myocardial function, as MW takes both deformation and afterload into account.
We also found that MWs in LAD segments had higher diagnostic value than MWs in LCX or RCA territories. In consistency with previous studies, the accuracy of 2D-STE was greater in the LAD territory, compared with other territories[22–24]. In our study, most MWs in LAD and RCA segments can predict relevant stenoses. However, in LCX segments, only MWI can predict LCX stenosis. Most MWs were found not able to identify stenosis in LCX. Therefore, results in these patients were analyzed in greater detail. One possible reason for lower predictive values of MWs in LCX or RCA territories could be inadequate visualization of adjacent myocardium in the inferior and posterior walls of left ventricular. Additionally, we found 70.8% severe stenoses (≥ 75% stenosis) in LCX were located in distal segment. Stenoses with more than 75% narrowing of the LAD and RCA coronary were mostly distributed in proximal-middle parts (84.8% and 71.4%). Preliminary studies showed that patients with proximal narrowings in any vessel had more severe ischemia and higher mortality rates than those with narrowings in distal locations in the same vessel[25, 26]. The diagnostic power of regional MWs may be closely related to the location of severe stenosis.