Moderated Posters session * Insights into the use of contrast stress echocardiography and 3D strain: 14/12/2013, 08:30-12:30 * Location: Moderated Poster area

Abstract

Purpose: Three-dimensional (3D) imaging allows for reproducible planes for analysis compared to two-dimensional (2D) imaging. We hypothesize that 3-D Dobutamine stress echocardiography (DSE) may offer higher diagnostic accuracy and prognostic value compared to 2-D. Methods: The 146 consecutive patients indicated forDSE(GE Vivid-7) had a standard 2D acquisition, longitudinal strain (LS) analysis by speckle tracking, 3D-triplane and full volume acquisitions with and without myocardial contrast (MCE, Sonovue, single shot 0.4 ml given at rest and peak stress) for left ventricular opacification. Chi-square test was done to assess the relationship between echo result and early revascularization. Kappa statistic was performed to assess agreement between 2D and 3D echo studies as well as between 3 independent readers with varying experience. Logistic regression analysis was done to predict late major cardiac events (late percutaneous intervention, CABG, myocardial infarction, cardiac death) at a follow up of 36 months. Results: The mean age was 63 ± 12 years (61% males) with 78/146 (53%) known coronary artery disease. Only 34% of the patients had excellent baseline image quality. Reliable LSassessment waspossible in85%at baseline but only in58%(p<0.0001) at peak stress despite maximizing frame rate. The 2D studies showed abnormal DSE in 42/146 patients, and 57/146 abnormal DSE with 3D MCE. The kappa statistic of agreement between 2D and 3D MCE was 0.106. For the 2D DSE, early revascularization was performed in 12/ 42(29%) of the abnormal studiesand16/104 (16%) of the normal studies, showing borderline significant relationship between 2DDSEresult and early revascularization (X2=3.25, df =1, p = 0.0712). While for the 3D-MCE, none of the normal studies (3D TomTec LVEF improved from 58% ± 6 to 74 ± 8% at peak stress, p < 0.001), and 28/57 (49%) of the abnormal studies (3D LVEF changed from 62 ± 5% to 58 ± 8% at peak stress, p=0.048) got early revascularization showing significant relationship between 3D-MCE DSEresultandearly revascularization (X2=50.47, df =1,p<0.0001. Therewasexcellent agreement between the 3 readers for DSE result (kappa=0.81). Logistic regression analysis showed a significant association between 3D-MCE DSEresult and late events (OR= 3.69 (95% CI 1.54 - 8.87), p=0.0035) but not between the 2DDSEresult and late events (p = 0.9908). Conclusion: 3-DMCEDSEcontributed toward improved diagnostic accuracy compared to 2-DDSEfor early revascularization, with excellent inter reader agreement. The probability of a late event was higher when 3D-MCE DSE was abnormal.