The Feasibility of Speckle Tracking for Intraoperative Assessment of Regional Myocardial Function by Transesophageal Echocardiography

Objectives: The authors aimed to examine the feasibility of intraoperative transesophageal echocardiography (TEE) acquisition of a non-Doppler-based, speckle tracking-derived myocardial deformation parameter (strain) immediately before and after coronary artery bypass graft (CABG) surgery in patients with reduced left ventricular (LV) function. Design: A clinical study. Setting: The cardiac surgery operating room of a tertiary referral institution. Patients: Ten patients with reduced LV function (ejection fraction lower than 35%) undergoing coronary revascularization were studied before and immediately after the procedure. Interventions: Perioperative TEE. Measurements and Results: A total of 120 myocardial segments were analyzed before and after CABG surgery. In visually obtained wall motion scoring (WMS), there were 29 normokinetic (N), 69 hypokinetic (H), 19 akinetic (A), and 3 dyskinetic (D) segments preoperatively and 26 N, 65 H, 21 A, and 8 D segments after CABG surgery. Preoperative radial strain correlated well with WMS (R = 0.82, p < 0.0001), whereas longitudinal strain showed only a weak correlation (R = 0.36, p < 0.0001). Postoperatively, correlations were similar. Interobserver variability as analyzed by kappa-statistics showed better agreement for radial (kappa = 0.82 +/- 0.05, p = 0.001) and longitudinal strain (kappa = 0.73 +/- 0.06, p = 0.004) than for WMS (kappa = 0.65 +/- 0.06). Preoperatively, strain was markedly greater in normally perfused segments than in ischemic segments, whereas the mean WMS revealed only minor differences. Conclusions: Strain calculation from TEE images is feasible during cardiac surgery and correlates well with WMS but has better interobserver agreement. Strain analysis, but not WMS, detected wall motion differences between normally perfused and ischemic segments. This simple method allows objective intraoperative quantification of myocardial segment function and may become an important monitoring tool in the future. (C) 2009 Elsevier Inc. All rights reserved.