Evaluation of Intraoperative Left-Ventricular Diastolic Function by Myocardial Strain in On-Pump Coronary Artery Bypass Surgery

Abstract

Left ventricular (LV) diastolic function strongly predicts outcomes after cardiac surgery, but there is no consensus about appropriate intraoperative assessment. Recently, intraoperative diastolic strain-based measurements assessed by transesophageal echocardiography (TEE) have shown a strong correlation with LV relaxation, compliance, and filling, but there are no reports about evaluation through the entire perioperative period. Therefore, the authors describe the intraoperative course of this novel assessment technique in patients who underwent coronary artery bypass grafting, and compare it with conventional echocardiographic measures and common grading algorithms of LV diastolic dysfunction (LVDD). Prospectively obtained data. A single university hospital. Thirty adult patients scheduled for isolated on-pump coronary artery bypass grafting surgery with preoperative preserved left and right ventricular systolic function, without significant heart valve disease and pulmonary hypertension, and an uneventful intraoperative course were included. Transesophageal echocardiography was performed after induction of anesthesia (T1), after termination of cardiopulmonary bypass (T2), and after sternal closure (T3). Echocardiographic evaluation was performed in stable hemodynamic conditions, in sinus rhythm or atrial pacing, and vasopressor support with norepinephrine ≤0.1 µg/kg/min. Strain-based measurements of peak longitudinal strain rate during isovolumetric relaxation (SR-IVR) and during early (SR-E) and late (SR-A) LV filling were assessed using EchoPAC v204 software (GE Vingmed Ultrasound AS, Norway). Evaluation of conventional echocardiographic parameters included transmitral Doppler measures of early (E) and late (A) LV filling, as well as lateral-tissue Doppler velocity assessed during early (e´) and late (a´) LV filling, tricuspid regurgitation, and left atrial dilatation. Evaluation and grading of LV diastolic function by myocardial strain was feasible in all included patients at all time points of assessment. Using conventional grading algorithms, however, a substantial number of patients could not be sufficiently graded, falling into an indeterminate zone and not reliably estimating LVDD (T1, 40%; T2, 33%; T3, 36%). There was significant impairment of LV diastolic function after bypass, as measured by SR-IVR (T1 v T2, 0.28 s-1 [IQR 0.23; 0.31) v 0.18 s-1 [IQR 0.14; 0.22]; p < 0.001), SR-E (T1 v T2, 0.95 ± 0.34 s-1v 1.28 ± 0.36 s-1; p < 0.001), and E/SR-IVR (T1 v T2, 2.3 ± 1.0 m v 4.5 ± 2.1 m; p < 0.001]. Conventional echocardiographic measures remained unchanged during the same period (E/A T1 v T2, 1.27 [IQR 0.94; 1.59] v 1.21 [IQR 1.03; 1.47] [p = 1] and E/e´ T1 v T2, 7.0 [IQR 5.3; 9.6] v 6.35 [IQR 5.7; 9.9] [p = 0.9]). There were no significant changes in the values of SR-IVR, SR-E, SR-A, E/SR-IVR, E/A, and E/e´ before and after sternal closure (T2 v T3). Intraoperative assessment of strain-based measurements of LV diastolic function and strain-based LVDD grading was feasible in this group of selected patients, whereas conventional parameters failed to describe LVDD sufficiently in a substantial number of patients. Diastolic strain-based measurements showed impairment of LV relaxation and compliance after bypass, which was not detected by conventional echocardiographic parameters. Therefore, diastolic myocardial strain analysis might be more sensitive in detecting myocardial diastolic dysfunction by TEE in the perioperative setting, with its dynamic changes of loading conditions, and might provide valuable and additional information on the perioperative changes of LV diastolic function.