The use of strain, strain rate, and displacement by 2D speckle tracking for assessment of systolic left ventricular function in goats: applicability and influence of general anesthesia

22 female Saanen goats aged 3.7 ± 1.1 [mean ± SD] years and with a body weight of 60.2 ± 10.5 kg were studied prospectively. All goats were considered healthy based upon physical examination, cardiac auscultation, and routine echocardiographic examination [1].

None of the goats received medications during the 2 weeks preceding entry into the study. The goats were acclimatized to the hospital for one week before the study. They were housed indoors, kept on straw and had free access to water and hay. Animal experiments were carried out in accordance with the Swiss law on animal protection. All animal studies were approved by the district veterinary office of the Canton of Zurich.

All goats underwent a complete echocardiographic examination while standing in a quiet room, unsedated and restrained by an experienced handler. A second echocardiographic examination within 7.4 ± 3.5 days of the first examination was performed immediately following a computed tomography scan conducted within the scope of another investigation, with the goats in general anesthesia and positioned in sternal recumbency. The mean time from induction of anesthesia to the start of the echocardiographic examination was 68 (52–103) min [mean (range)].

The animals were fasted for 24 h and deprived of water for 2 h prior to anaesthesia. A 14 G/1.88 in catheter (BD Angiocath, Becton Dickinson, Allschwil, Switzerland) was placed aseptically into the right jugular vein. All animals were premedicated intravenously with 0.1 mg/kg xylazine (Rompun 2%, Provet, Lyssach, Switzerland), diluted in 20 mL of saline (NaCl 0.9%, Braun Medical, Sempach, Switzerland) and given over 5 min with a syringe pump (Syramed μ6000, Arcomed, Regensdorf, Switzerland). Anaesthesia was then induced with either 3 mg/kg racemic ketamine (Narketan 10%, Vetoquinol, Ittigen, Switzerland) or 1.5 mg/kg S-ketamine (Keta-S 6%, Dr. Graeub AG, Bern, Switzerland) i.v. given by manual injection.

Once the animals were recumbent, they were positioned sternally, intubated, and connected to an anesthesia machine. Generally, the goats were allowed to breathe spontaneously throughout the entire duration of anesthesia. Mechanical ventilation was only applied when arterial blood gas analysis revealed a PaCO₂ > 50 mmHg. In this case, intermittent positive pressure ventilation was used with a tidal volume of 10-15 mL/kg at a rate of 10-15 breaths per minute. Anesthesia was maintained with isoflurane (IsoFlo®, Abbott, Baar, Switzerland) delivered in oxygen and air via a semi-closed circle absorption system with a flow of 60 mL/kg/min during the first 10 min and a flow of 35 mL/kg/min thereafter. The vaporizer was initially set to 2.5% and after 15 min it was reduced and held between 1.5 and 2% to maintain an end-tidal isoflurane concentration of 1.1%. Lactated Ringer’s solution (Ringer-Laktat, Fresenius Kabi, Stans, Switzerland) was administered i.v. at a dose rate of 10 mL/kg/h. During anesthesia, the goats were kept in sternal recumbency.

Each goat was instrumented with an electrocardiogram, a pulse oximeter and an end-tidal gas analyzer. A portable multiparameter monitor (Datex-Ohmeda AS/3 Compact Monitor; Anandic, Schaffhausen, Switzerland) continuously displayed the physiologic monitoring data during the anesthetic period. The ear artery was cannulated for intraarterial blood pressure monitoring and blood sampling for blood gas analyses. Heart rate (HR), respiratory rate, saturation of arterial oxygen, direct arterial blood pressures, end-tidal CO₂, end-tidal isoflurane, and rectal temperature were monitored. Blood gas analysis was performed at 10 minutes after anesthesia induction and then every 30 minutes for the duration of anesthesia. Following anesthesia, the goats were placed sternally in a padded box. The endotracheal tube was removed once swallowing reflex returned.

Transthoracic echocardiography was performed using a high-end digital echocardiograph (GE Vivid 7 Dimension, BTO6, GE Medical Systems, Glattbrugg, Switzerland) with a phased array transducer (M4S, GE Medical Systems, Glattbrugg, Switzerland) at a frequency of 1.9/4.0 MHz (octave harmonics) [22]. A single lead electrocardiogram (lead I) was recorded simultaneously. The lead position was consistent among animals and between repeat echocardiograms. The aortic valve was imaged in 2D grayscale mode in a right parasternal long-axis view of the left ventricular outflow tract (LVOT) and in M-mode in a right parasternal short-axis view, respectively, for subsequent determination of the time of aortic valve closure. The left ventricle (LV) was imaged in 2D mode using a right parasternal four-chamber view optimized to obtain an image of the entire LV at its largest dimensions, including the LV apex. Subsequently, the LV was imaged in three short-axis views, namely at the level of the apex, at the level of the papillary muscles, and at the level of the chordae tendineae. Imaging depth and sector width, respectively, were adjusted to achieve a frame rate > 50 frames/second in 2D imaging mode. Three representative, non-consecutive cardiac cycles were recorded in each view and stored as cine-loops in digital raw data format. Data analyses were performed offline, blinded, and in random order, using a dedicated software package (EchoPAC Software version 6.1.2, GE Medical Systems, Glattbrugg, Switzerland). Three cardiac cycles were analyzed for each imaging plane. For each measured or calculated variable, the average of the three measurements was reported.

The time to aortic valve closure was measured manually (tAVC_m). It was defined as the time interval between the peak of the electrocardiographic R wave and the closure point of the aortic valve identified on an M-mode recording of aortic valve motion. In cases, in which the closure point of the aortic valve was not clearly identifiable on the M-mode recording, anatomical M-mode was applied to the recorded two-dimensional echo loop of the LVOT, and cursor placement was adjusted to identify the point of aortic valve closure. The corresponding heart rate (HR_AVCm) for each cycle was calculated as 60,000/R-R interval.

The 2D speckle tracking (2DST) analyses were performed using the 2D Strain module of the analysis software (EchoPAC Software version 6.1.2, GE Medical Systems, Glattbrugg, Switzerland). The 2DST variables were measured as follows: (1) The appropriate long-axis or short-axis image was selected and the Q-Analysis module was started. (2) A single heart cycle was selected by moving the left and the right cursor, respectively, to the peak of the R waves of the electrocardiogram. Each cycle was identified by the image number, the time when the image was recorded, and the first and the last frame of the cycle. The R-R interval was measured and the frame rate was recorded. (3) Subsequently, the 2D Strain module was started. The long-axis (subsequently called LAX) grayscale loops of the LV were analyzed using the “4CH” option, and the short-axis grayscale loops of the LV were analyzed using the “SAX-AP” (apical level, subsequently called SAX-AP), “SAX-PM” (papillary muscle level, subsequently called SAX-PM), and “SAX-MV” option (chordal level, subsequently called SAX-CH), respectively. Once the option had been selected, a region of interest (ROI) was determined by tracing the endocardial border of the LV at end-systole. For the long-axis images, tracing started at the septal mitral valve (MV) annulus and ended at the lateral MV annulus; for the short-axis images, tracing was started at mid-septum and proceeded in a clockwise direction. The papillary muscles were not included in the tracings. The ROI width was adjusted, so that the entire myocardial thickness was covered throughout the cardiac cycle. Subsequently the speckle tracking analysis was started. The software algorithm automatically divided the myocardium into 6 segments, performed the speckle tracking analysis, and provided confirmation of adequate tracking for each segment (Additional file 1: Figure S5). The segments were preselected by the software based on regional wall motion analysis standards applied to human patients and were not adjusted for use in the goat. Hence, apart from slight individual variations, the short-axis segments termed “Sept”, “AntSept” and “Ant” generally depict the interventricular septum, whereas “Lat”, “Post”, and “Inf” depict the cranial, lateral, and caudal LV free wall.

The quality of the tracking was visually assessed by the operator during motion playback. If necessary, the line tracing of the endocardium was adjusted and the speckle tracking analysis was repeated until adequate tracking was achieved. If adequate tracking was not possible despite repeated adjustments of the ROI, another cardiac cycle was chosen for analysis. If the software failed to adequately track a segment even after repeated manual tracing of multiple cycles, the loop was excluded from analysis.

Six curve profiles were obtained corresponding to the average of each myocardial segment (Additional file 1: Figure S6). For the long-axis images (4CH), the following variables were reported by the software: Longitudinal strain (ε_L), longitudinal strain rate (SR_L), longitudinal displacement (D_L), and transverse displacement (D_T). For the short-axis images (SAX-AP, SAX-PM, SAX-CH), the following variables were reported by the software: Circumferential strain (ε_C), circumferential strain rate (SR_C), radial strain (ε_R), radial strain rate (SR_R), radial displacement (D_R). The time of aortic valve closure was automatically calculated and displayed by the software (AVC_a, Additional file 1: Figure S6).

The measurements for strain, strain rate, and displacement, respectively, were performed on the “Results” screen of the 2D Strain software module. Automated detection of the peak values was verified on the graphical display and corrected as necessary. The mean of the 6 segmental measurements of each variable was calculated to obtain indices of averaged strain, strain rate, and displacement. Strain measurements for each segment included longitudinal, radial, and circumferential peak strain (peak independent of aortic valve closure; termed ε_L, ε_R, and ε_C) as well as peak systolic strain (peak prior to or at the time of aortic valve closure; termed ε_L-sys, ε_R-sys, and ε_C-sys). Where the highest strain value occurred before or at the time of aortic valve closure, peak strain (ε) was identical with peak systolic strain (ε_sys). Where 2 peaks for strain were present and ε_sys was followed by a 2^nd, higher peak after aortic valve closure, the 2^nd peak was considered ε. Where only 1 peak occurring after aortic valve closure was present, ε_sys was defined as strain at the time of aortic valve closure and the 2^nd peak was termed ε. Post-systolic motion was diagnosed where the highest strain value occurred after automatically determined aortic valve closure (AVC_a) [7,9,13,14].

Strain rate measurements for each segment included peak systolic strain rate (SR_L-sys, SR_R-sys, and SR_C-sys), peak early-diastolic strain rate (SR_L-E, SR_R-E, and SR_C-E) and peak late-diastolic strain rate (SR_L-A, SR_R-A, and SR_C-A). Displacement measurements for each segment included peak displacement (D_L, D_T, D_R). The R-R interval was measured for each cycle. The time to automatically determined aortic valve closure (tAVC_a) was calculated as the time interval between the peak of the electrocardiographic R wave and AVC_a displayed on the “Trace” screen. The corresponding heart rate (HR_AVCa) for each cycle was calculated as 60,000/R-R interval.

The time interval from the electrocardiographic R wave to longitudinal, circumferential, and radial peak strain (t_εL, t_εC, t_εR) of each segment was measured in each cycle. The synchrony time index (STI_ε), a measure of myocardial dyssynchrony, was calculated as the difference in t_ε from the earliest to the latest segment [12,13].

All statistical and graphical analyses were performed using standard computer software (Microsoft Office Excel 2003, Microsoft Corporation, Redmond, WA; SigmaStat v3.5, SPSS Inc, Chicago, IL; GraphPad Prism v5.00 for Windows, GraphPad Software, San Diego California USA). The agreement between tAVC_a and tAVC_m was compared by paired t test and Bland-Altman statistics. The corresponding heart rates (HR_AVCm and HR_AVCa) were compared by paired t test. For segmental 2DST indices, 2-way repeated-measures analysis of variance was used to detect differences between segments and treatment (i.e., awake vs. anesthetized). When the F test indicated significant differences, all pairwise multiple comparisons were performed using the Holm-Sidak post hoc test. The effect of general anesthesia on averaged 2DST indices and on STI_ε was assessed by paired t tests, reporting the 95% confidence intervals for the difference of means. Validity of the normality assumption was confirmed by assessment of normal probability plots of the residuals. The level of significance was set at P=0.05.