Background: The potential advantage of real time 3D stress echocardiography is to acquire the same cycle for parasternal and apical views at the expense of a lower frame rate, compared to the conventional 2D approach. We aimed at evaluating the feasibility of real time three dimensional echocardiography (3D) during dobutamine stress echocardiography (DSE) in comparison with conventional two dimensional approach (2D).

Methods: We acquired, using a commercially available machine (VIVID7, equipped with an ECHOPAC software package) during the same session both 2D and 3D DSE in 50 patients (mean age 62 € 11 years, 33 men) referred for evaluation of ischemia in known or suspected coronary artery disease (CAD). Dobutamine was infused using a standard protocol with atropine if necessary to reach the target heart rate. Images acquisition was performed according to the following protocol: 2D images were first obtained, then 3D at each stage. We aimed at determining the agreement in segment visualization for each wall and all stages. We also measured the long axis diameter of the LV in all apical views at rest with both approaches. All examinations were reviewed by the same investigator; 2D and 3D images for each patient were not analysed during the same session. The agreement between the results (dichotomised as interpretable and non interpretable) obtained by 2D and 3D approaches was determined using Kappa-correlation coefficient for segmental wall motion analysis and Pearson test for LV long-axis diameter measurements.

Results: Prior myocardial infarction was documented in 8% of patients and 14% of patients had prior revascularisation. The acquisition was completed successfull with both methods (2D and 3D) in all patients. Mean ? SD heart rate at baseline and peak stress was 74.10 ? 14.73 and 149.42 ? 11.78 bpm respectively. Systolic blood pressure was 137.6 ? 23.29 and 148.5 ? 33.61 mmHg, respectively.

In a total of 800 segments, the number of interpretable segments acquired by the 2D segments was 791 (98.8%) and that by the RT3D was 770 (96%) p = 0.001). Using a segment to segment comparison, we found an agreement varying from 92% to 100%. Interobserver agreement, performed in 5 examinations, for visualisation of interpretable segments was 99%. Pearson correlation for LV long – axis measurement from apical 4C, -3C, -2C was 0.80 (p < 0.001), 0.83 (p < 0.001), and 0.81 (p < 0.001), respectively.

Conclusion: In this preliminary study, aiming at learning how to use 3D approach during DSE, real time 3D DSE seems feasible compared with conventional 2D imaging, despite a lower frame rate. A larger trial including coronary angiography is needed to confirm the feasibility and to establish the efficacy and accuracy in the detection of ischemia and also to test the hypothesis that 3D might help to save time during acquisition.