© 2015, Wiley Periodicals, Inc. DOI: 10.1111/echo.12983
Echocardiography
A New Method for Direct Three-Dimensional Measurement of Left Atrial Appendage Dimensions during Transesophageal Echocardiography Chaim Yosefy, M.D.,* Avishag Laish-Farkash, M.D., Ph.D.,* Yulia Azhibekov, M.D.,† Vladimir Khalameizer, M.D.,* Boris Brodkin, M.D.,* and Amos Katz, M.D.* *Department of Cardiology, Barzilai Medical Center, Ben-Gurion University of the Negev, Ashkelon, Israel; and †Department of Imaging, Barzilai Medical Center, Ben-Gurion University of the Negev, Ashkelon, Israel
Aims: Currently, two-dimensional transesophageal echocardiography (2DTEE) at a cut-plane angulation of 135° is the recommended method to size maximal left atrial appendage (LAA) orifice diameter before introducing a percutaneous LAA closure device. We compared real time three-dimensional TEE (RT3DTEE) and 2DTEE for measuring LAA dimensions versus computed tomography (CT) as gold standard. Methods and Results: We prospectively studied 30 consecutive patients who underwent a routine TEE examination, using QLAB 10.0 Application on EPIQ7 iE33 3D echo machine between December 2012 and December 2013. All patients underwent 64-slice CT before pulmonary vein isolation or for workup of pulmonary embolism. LAA measurements were compared between 135 2DTEE and RT3DTEE. Results were compared with CT measurements. Using RT3DTEE, larger LAA diameters were measured versus 2DTEE (23.5 3.9 vs. 24.5 4.7 mm). In seven patients (23.3%), the measurements in 135° 2DTEE were smaller than the cut-plane angulation with maximal orifice diameter. RT3DTEE measurements of LAA were not different from CT regarding number of lobes, area of orifice, and maximal diameter. LAA volume could not be measured directly using RT3DTEE. No difference was found between LAA depth using RT3DTEE (19.5 2.3 mm) vs. CT (19.6 2.3, P = NS) and 2DTEE (19.4 2.2 mm) vs. CT (P = NS). However, RT3DTEE (24.5 4.7 mm) vs. CT (24.6 5, P = NS) was more accurate in measuring maximal LAA diameter compared to 2DTEE (23.5 3.9 mm) vs. CT (P < 0.01). Conclusion: RT3DTEE method is more accurate than 2DTEE for assessment of maximal LAA orifice diameter. Bedsides, RT3DTEE LAA measurements are not statistically different from CT. Thus, RT3DTEE may facilitate LAA closure procedure by choosing the appropriate device size. (Echocardiography 2016;33:69–76) The left atrial appendage (LAA) is considered the main site of thrombus formation, inducing stroke in patients suffering from nonvalvular atrial fibrillation (AF).1 Low stroke rates are reported in patients in whom the LAA has been surgically removed,2–4 but there is a lack of large, controlled trials with systematic follow-up. More recently, minimally invasive epicardial techniques and catheter-based transseptal techniques have been developed for occlusion of the LAA orifice to reduce stroke risk.5–7 These devices and procedures may provide an alternative to oral anticoagulation (OAC) for AF patients at high risk for stroke but with contraindications for chronic OAC.8 Currently, two self-expanding devices, the WATCHMAN (Boston Scientific, Natick, MA, USA) The first two authors contributed equally to the article. Address for correspondence and reprint requests: Chaim Yosefy, Noninvasive Cardiology Unit, Barzilai Medical Center, Ashkelon 78306, Israel. Fax: +972-8-674-567; E-mail: [email protected]
and the Amplatzer Cardiac Plug (St. Jude Medical, Inc., St. Paul, MN, USA), which are transseptally placed in the LAA, are available for clinical use in Europe, while their evaluation in controlled trials is still in progress.9 Precise knowledge of LAA anatomy and dimensions has become a key guiding stage before introducing the LAA closure devices.10,11 Complex LAA morphology characterized by an increased number of LAA lobes was associated with the presence of LAA thrombus12 and should be ruled out prior to procedure.13–15 Currently, two-dimensional transesophageal echocardiography (2DTEE) at an angle of 135 degrees is the recommended method to size maximal LAA orifice diameter.10,11,16,17 However, 2DTEE does not adequately allow complete spatial visualization of the LAA.13–15 Consequently, threedimensional imaging modalities,13–15 such as real time three-dimensional transesophageal echocardiography (RT3DTEE), cardiac magnetic reso-
69
Yosefy, et al.
nance imaging (MRI), and computed tomography (CT), may be more accurate.17–20 Recent evolving data show better performance of RT3DTEE for the assessment of LAA anatomy compared with 2DTEE regarding LAA orifice area, LAA ejection fraction calculation, and LAA volume.17,21–23 The aim of our study was twofold: First, to compare the RT3DTEE measurements of LAA diameter with those obtained with conventional 2DTEE, 64-slice CT served as the reference technique; second, to investigate whether sizing LAA at 135° cut-plane angulation with 2DTEE, as currently recommended, is indeed the optimal angulation that enables measuring the maximal LAA diameter. If this is not the case, then RT3DTEE may have an advantage over 2DTEE, being rotational and nondependent on cut-plane angulation. Accuracy in the LAA diameter measurement using RT3DTEE may facilitate LAA closure procedure by choosing the appropriate closing device size. Methods: Study Population: A total of 30 consecutive patients who underwent a routine indicated 2DTTE (two-dimensional transthoracic echocardiography), 2DTEE, RT3DTEE, and 64-slice CT, either before pulmonary vein isolation (PVI) ablation (n = 23) for precise definition of LA and pulmonary veins anatomy, or for workup of pulmonary embolism (PE) (n = 7) (Table I). In this group of patients, 64-slice CT was used as reference technique to test the accuracy of 2DTEE- and RT3DTEE-derived measurements of LAA parameters. The study protocol was approved by our institution Helsinki review board and was not supported by any company. Echocardiography: Thirty consecutive patients (of 34 patients) who underwent the routine echocardiography examination using iE33 echo machine (Philips Medical Systems, Andover, MA) between December 2012 and December 2013 and had a good echogenic window, were included in the study. All images were digitally stored for offline analysis (QLAB 10.0 cardiac 3DQ, Philips Medical Systems). Area of LAA orifice, LAA depth, maximal LAA diameter, number of LAA lobes, and LAA volume were compared between RT3DTEE and 2DTEE to the CT as the gold standard method. All echocardiographic data have been reviewed by a single operator (CY) who was blinded to CT results performed by a single operator (YA).
70
TABLE I Baseline Demographics and Echocardiographic Characteristics of the Study Population (n = 30) Study Population (n = 30) Age (years) Male (%) BSA (m2) Height (meter) Weight (Kg) Indication of CT Echocardiographic measurements (RT3DTEE) LVEF (est. %) LPW-D (mm) IVS-D (mm) LVESD (mm) LVEDD (mm) Pulmonary pressure (mmHg) RA area (cm2) LA area (cm2) LA diameter (ap) (mm) Ascending aorta diameter (mm) Aortic root diameter (mm)
64.3 11.8 16 (53) 1.9 0.2 1.7 0.1 83 16.7 Before PVI - 23 PE workup - 7
59.3 10 10.7 31.6 49.2 34.5 18.7 22.9 40.1 32.6 30.3
7.7 2 2.2 3.7 4.2 15.1 4.4 4.7 6.6 3.6 3.1
BSA = body surface area; CT = computed tomography; est = estimated; IVS-D = interventricular septum diameter; LPW-D = left posterior ventricular wall diameter; LVEDD = left ventricular end-diastolic diameter; LVEF = left ventricular ejection fraction; LVESD = left ventricular end-systolic diameter; PE = pulmonary embolism; PVI = pulmonary vein isolation; RT3DTEE = real time 3-dimensional transesophageal echocardiography.
2DTTE: 2DTTE was performed using the S5-1-MHz transducer. Standard M-mode, 2D images, and Doppler and color Doppler data were acquired from the parasternal and apical views (4-, 2-, and 3-chamber) and digitally stored in cine-loop format. Transesophageal Echocardiography: Transesophageal echocardiography was performed using commercially available fully sampled matrix-array TEE transducer and ultrasound system (X7–2t Live 3D TEE transducer). 2DTEE: 2DTEE was performed according to a standard clinical protocol.24 The LAA was imaged at 0°, 45°, 90°, and 135° (Fig. 1). The 2D images were analyzed online. Measurements of maximum (D1) and minimum (D2) diameters of LAA orifice were obtained from orthogonal planes (0° and 90°, and 45° and 135°), from the origin of the left circumflex artery to the roof of the LAA, below the ligament of Marshall. Similarly, the LAA depth (i.e., the distance from LAA orifice to the tip of LAA) was measured at 0°, 45°, 90°, and 135°; from these four views, the longest LAA depth was then chosen.22
Direct 3D Measurement of Left Atrial Appendage in TEE
Figure 1. Two-dimensional transesophageal echocardiography (2DTEE) measurements of left atrial appendage (LAA) diameter (D) and depth at 0°, 45°, 90°, and 135° cut-plane angulation.
RT3DTEE: RT3DTEE imaging was performed acquiring a pyramidal data set large enough to include the entire LAA, using the zoom mode. Subsequently, pyramidal data sets were cropped along designated x-, y-, and z-axes or using an arbitrary cropping plane, to remove remaining nonrelevant anatomic structures and to improve the visualization of LAA. The LAA depth (i.e., the longest distance from LAA orifice to the tip of LAA) was measured offline from the long-axes views, using dedicated software (QLAb 10.0). On these data sets, we tried to measure LAA 3D volume using the same tracking method that we used for LA volume measurement, as previously described.25 The area of the LAA orifice, as well as the maximum (D1) and minimum (D2) diameters of LAA orifice, was measured directly from the original 3D views, along a plane connecting the origin of the left circumflex artery to the roof of the LAA, below the ligament of Marshall (Fig. 2). These measurements were assessed online using the EPIQ7 echo machine (Philips Medical Systems), because they could not be measured offline on QLAB 10.0 software. This method is different from the measurements assessed using the multiplanar reconstruction (MPR) mode in the shortaxis view described elsewhere.21 All 2D and 3D measurements of LAA were performed at ventricular end-systole. In patients who were in atrial fibrillation at the time of TTE and TEE, measurements from ≥3 cardiac cycles were averaged.
Sixty-Four-Slice Computed Tomography: All thirty patients underwent clinically indicated 64-slice CT (Philips Brilliance CT 64 Power-Philips Medical Systems, Eindhoven, The Netherlands) within one week of transthoracic and transesophageal echocardiography. All patients were in sinus rhythm during the CT scan. A retrospective ECGgating protocol was used. Scanning parameters were the following: detector collimation of 0.625 mm, total z-axis coverage of 40 mm per rotation, gantry rotation speed of 0.35 seconds, tube voltage of 120 kV, pitch of 0.16 to 0.24, and EKG modulated tube current ranging from 400 to 800 mA. The bolus tracking technique was used to trigger the acquisition, with a four-cavity view as the region of interest. A total of 70–100 mL of iodinated, nonionic contrast agent (Iomeron 350, Bracco Imaging S.p.A., Milan, Italy) was injected continuously into the antecubital vein (100– 120 mL at 5.0 mL/sec). Scanning was initiated during a single breath hold for an acquisition time of 5 to 7 seconds. All images were reconstructed with an effective slice thickness of 0.625 mm. ECG-gating protocol reconstruction of the image data was performed starting from early systole (10% of RR interval) and ending at end-diastole (90% of RR interval) using 10% steps. Reconstructed image data were transferred to a remote workstation (IntelliSpace Portal, Philips) for postprocessing. For the purpose of this study, image data sets reconstructed at end-systole (40% of RR interval) were used for analysis. Using MPR, measurements of the area of the LAA orifice were performed from the 71
Yosefy, et al.
Figure 2. Direct measurement in 360° rotating three-dimensional real time transesophageal echocardiography (RT3DTEE) of left atrial appendage (LAA) orifice area at the level of circumflex (Cx) artery (arrow). D = largest diameter.
short-axis view as well as the maximum (D2) diameters, and the maximum (D1) depth of LAA (the longest distance from LAA orifice to the tip of LAA and orifice) (Fig. 3). Statistical Analysis: Continuous variables are presented as percentages and means standard deviation. Categorical data are presented as absolute numbers and percentages. Continuous variables were compared using independent Student’s t-test; categorical variables were compared using chisquare test or Fisher’s exact test. A 2-sided Pvalue
Echocardiography
A New Method for Direct Three-Dimensional Measurement of Left Atrial Appendage Dimensions during Transesophageal Echocardiography Chaim Yosefy, M.D.,* Avishag Laish-Farkash, M.D., Ph.D.,* Yulia Azhibekov, M.D.,† Vladimir Khalameizer, M.D.,* Boris Brodkin, M.D.,* and Amos Katz, M.D.* *Department of Cardiology, Barzilai Medical Center, Ben-Gurion University of the Negev, Ashkelon, Israel; and †Department of Imaging, Barzilai Medical Center, Ben-Gurion University of the Negev, Ashkelon, Israel
Aims: Currently, two-dimensional transesophageal echocardiography (2DTEE) at a cut-plane angulation of 135° is the recommended method to size maximal left atrial appendage (LAA) orifice diameter before introducing a percutaneous LAA closure device. We compared real time three-dimensional TEE (RT3DTEE) and 2DTEE for measuring LAA dimensions versus computed tomography (CT) as gold standard. Methods and Results: We prospectively studied 30 consecutive patients who underwent a routine TEE examination, using QLAB 10.0 Application on EPIQ7 iE33 3D echo machine between December 2012 and December 2013. All patients underwent 64-slice CT before pulmonary vein isolation or for workup of pulmonary embolism. LAA measurements were compared between 135 2DTEE and RT3DTEE. Results were compared with CT measurements. Using RT3DTEE, larger LAA diameters were measured versus 2DTEE (23.5 3.9 vs. 24.5 4.7 mm). In seven patients (23.3%), the measurements in 135° 2DTEE were smaller than the cut-plane angulation with maximal orifice diameter. RT3DTEE measurements of LAA were not different from CT regarding number of lobes, area of orifice, and maximal diameter. LAA volume could not be measured directly using RT3DTEE. No difference was found between LAA depth using RT3DTEE (19.5 2.3 mm) vs. CT (19.6 2.3, P = NS) and 2DTEE (19.4 2.2 mm) vs. CT (P = NS). However, RT3DTEE (24.5 4.7 mm) vs. CT (24.6 5, P = NS) was more accurate in measuring maximal LAA diameter compared to 2DTEE (23.5 3.9 mm) vs. CT (P < 0.01). Conclusion: RT3DTEE method is more accurate than 2DTEE for assessment of maximal LAA orifice diameter. Bedsides, RT3DTEE LAA measurements are not statistically different from CT. Thus, RT3DTEE may facilitate LAA closure procedure by choosing the appropriate device size. (Echocardiography 2016;33:69–76) The left atrial appendage (LAA) is considered the main site of thrombus formation, inducing stroke in patients suffering from nonvalvular atrial fibrillation (AF).1 Low stroke rates are reported in patients in whom the LAA has been surgically removed,2–4 but there is a lack of large, controlled trials with systematic follow-up. More recently, minimally invasive epicardial techniques and catheter-based transseptal techniques have been developed for occlusion of the LAA orifice to reduce stroke risk.5–7 These devices and procedures may provide an alternative to oral anticoagulation (OAC) for AF patients at high risk for stroke but with contraindications for chronic OAC.8 Currently, two self-expanding devices, the WATCHMAN (Boston Scientific, Natick, MA, USA) The first two authors contributed equally to the article. Address for correspondence and reprint requests: Chaim Yosefy, Noninvasive Cardiology Unit, Barzilai Medical Center, Ashkelon 78306, Israel. Fax: +972-8-674-567; E-mail: [email protected]
and the Amplatzer Cardiac Plug (St. Jude Medical, Inc., St. Paul, MN, USA), which are transseptally placed in the LAA, are available for clinical use in Europe, while their evaluation in controlled trials is still in progress.9 Precise knowledge of LAA anatomy and dimensions has become a key guiding stage before introducing the LAA closure devices.10,11 Complex LAA morphology characterized by an increased number of LAA lobes was associated with the presence of LAA thrombus12 and should be ruled out prior to procedure.13–15 Currently, two-dimensional transesophageal echocardiography (2DTEE) at an angle of 135 degrees is the recommended method to size maximal LAA orifice diameter.10,11,16,17 However, 2DTEE does not adequately allow complete spatial visualization of the LAA.13–15 Consequently, threedimensional imaging modalities,13–15 such as real time three-dimensional transesophageal echocardiography (RT3DTEE), cardiac magnetic reso-
69
Yosefy, et al.
nance imaging (MRI), and computed tomography (CT), may be more accurate.17–20 Recent evolving data show better performance of RT3DTEE for the assessment of LAA anatomy compared with 2DTEE regarding LAA orifice area, LAA ejection fraction calculation, and LAA volume.17,21–23 The aim of our study was twofold: First, to compare the RT3DTEE measurements of LAA diameter with those obtained with conventional 2DTEE, 64-slice CT served as the reference technique; second, to investigate whether sizing LAA at 135° cut-plane angulation with 2DTEE, as currently recommended, is indeed the optimal angulation that enables measuring the maximal LAA diameter. If this is not the case, then RT3DTEE may have an advantage over 2DTEE, being rotational and nondependent on cut-plane angulation. Accuracy in the LAA diameter measurement using RT3DTEE may facilitate LAA closure procedure by choosing the appropriate closing device size. Methods: Study Population: A total of 30 consecutive patients who underwent a routine indicated 2DTTE (two-dimensional transthoracic echocardiography), 2DTEE, RT3DTEE, and 64-slice CT, either before pulmonary vein isolation (PVI) ablation (n = 23) for precise definition of LA and pulmonary veins anatomy, or for workup of pulmonary embolism (PE) (n = 7) (Table I). In this group of patients, 64-slice CT was used as reference technique to test the accuracy of 2DTEE- and RT3DTEE-derived measurements of LAA parameters. The study protocol was approved by our institution Helsinki review board and was not supported by any company. Echocardiography: Thirty consecutive patients (of 34 patients) who underwent the routine echocardiography examination using iE33 echo machine (Philips Medical Systems, Andover, MA) between December 2012 and December 2013 and had a good echogenic window, were included in the study. All images were digitally stored for offline analysis (QLAB 10.0 cardiac 3DQ, Philips Medical Systems). Area of LAA orifice, LAA depth, maximal LAA diameter, number of LAA lobes, and LAA volume were compared between RT3DTEE and 2DTEE to the CT as the gold standard method. All echocardiographic data have been reviewed by a single operator (CY) who was blinded to CT results performed by a single operator (YA).
70
TABLE I Baseline Demographics and Echocardiographic Characteristics of the Study Population (n = 30) Study Population (n = 30) Age (years) Male (%) BSA (m2) Height (meter) Weight (Kg) Indication of CT Echocardiographic measurements (RT3DTEE) LVEF (est. %) LPW-D (mm) IVS-D (mm) LVESD (mm) LVEDD (mm) Pulmonary pressure (mmHg) RA area (cm2) LA area (cm2) LA diameter (ap) (mm) Ascending aorta diameter (mm) Aortic root diameter (mm)
64.3 11.8 16 (53) 1.9 0.2 1.7 0.1 83 16.7 Before PVI - 23 PE workup - 7
59.3 10 10.7 31.6 49.2 34.5 18.7 22.9 40.1 32.6 30.3
7.7 2 2.2 3.7 4.2 15.1 4.4 4.7 6.6 3.6 3.1
BSA = body surface area; CT = computed tomography; est = estimated; IVS-D = interventricular septum diameter; LPW-D = left posterior ventricular wall diameter; LVEDD = left ventricular end-diastolic diameter; LVEF = left ventricular ejection fraction; LVESD = left ventricular end-systolic diameter; PE = pulmonary embolism; PVI = pulmonary vein isolation; RT3DTEE = real time 3-dimensional transesophageal echocardiography.
2DTTE: 2DTTE was performed using the S5-1-MHz transducer. Standard M-mode, 2D images, and Doppler and color Doppler data were acquired from the parasternal and apical views (4-, 2-, and 3-chamber) and digitally stored in cine-loop format. Transesophageal Echocardiography: Transesophageal echocardiography was performed using commercially available fully sampled matrix-array TEE transducer and ultrasound system (X7–2t Live 3D TEE transducer). 2DTEE: 2DTEE was performed according to a standard clinical protocol.24 The LAA was imaged at 0°, 45°, 90°, and 135° (Fig. 1). The 2D images were analyzed online. Measurements of maximum (D1) and minimum (D2) diameters of LAA orifice were obtained from orthogonal planes (0° and 90°, and 45° and 135°), from the origin of the left circumflex artery to the roof of the LAA, below the ligament of Marshall. Similarly, the LAA depth (i.e., the distance from LAA orifice to the tip of LAA) was measured at 0°, 45°, 90°, and 135°; from these four views, the longest LAA depth was then chosen.22
Direct 3D Measurement of Left Atrial Appendage in TEE
Figure 1. Two-dimensional transesophageal echocardiography (2DTEE) measurements of left atrial appendage (LAA) diameter (D) and depth at 0°, 45°, 90°, and 135° cut-plane angulation.
RT3DTEE: RT3DTEE imaging was performed acquiring a pyramidal data set large enough to include the entire LAA, using the zoom mode. Subsequently, pyramidal data sets were cropped along designated x-, y-, and z-axes or using an arbitrary cropping plane, to remove remaining nonrelevant anatomic structures and to improve the visualization of LAA. The LAA depth (i.e., the longest distance from LAA orifice to the tip of LAA) was measured offline from the long-axes views, using dedicated software (QLAb 10.0). On these data sets, we tried to measure LAA 3D volume using the same tracking method that we used for LA volume measurement, as previously described.25 The area of the LAA orifice, as well as the maximum (D1) and minimum (D2) diameters of LAA orifice, was measured directly from the original 3D views, along a plane connecting the origin of the left circumflex artery to the roof of the LAA, below the ligament of Marshall (Fig. 2). These measurements were assessed online using the EPIQ7 echo machine (Philips Medical Systems), because they could not be measured offline on QLAB 10.0 software. This method is different from the measurements assessed using the multiplanar reconstruction (MPR) mode in the shortaxis view described elsewhere.21 All 2D and 3D measurements of LAA were performed at ventricular end-systole. In patients who were in atrial fibrillation at the time of TTE and TEE, measurements from ≥3 cardiac cycles were averaged.
Sixty-Four-Slice Computed Tomography: All thirty patients underwent clinically indicated 64-slice CT (Philips Brilliance CT 64 Power-Philips Medical Systems, Eindhoven, The Netherlands) within one week of transthoracic and transesophageal echocardiography. All patients were in sinus rhythm during the CT scan. A retrospective ECGgating protocol was used. Scanning parameters were the following: detector collimation of 0.625 mm, total z-axis coverage of 40 mm per rotation, gantry rotation speed of 0.35 seconds, tube voltage of 120 kV, pitch of 0.16 to 0.24, and EKG modulated tube current ranging from 400 to 800 mA. The bolus tracking technique was used to trigger the acquisition, with a four-cavity view as the region of interest. A total of 70–100 mL of iodinated, nonionic contrast agent (Iomeron 350, Bracco Imaging S.p.A., Milan, Italy) was injected continuously into the antecubital vein (100– 120 mL at 5.0 mL/sec). Scanning was initiated during a single breath hold for an acquisition time of 5 to 7 seconds. All images were reconstructed with an effective slice thickness of 0.625 mm. ECG-gating protocol reconstruction of the image data was performed starting from early systole (10% of RR interval) and ending at end-diastole (90% of RR interval) using 10% steps. Reconstructed image data were transferred to a remote workstation (IntelliSpace Portal, Philips) for postprocessing. For the purpose of this study, image data sets reconstructed at end-systole (40% of RR interval) were used for analysis. Using MPR, measurements of the area of the LAA orifice were performed from the 71
Yosefy, et al.
Figure 2. Direct measurement in 360° rotating three-dimensional real time transesophageal echocardiography (RT3DTEE) of left atrial appendage (LAA) orifice area at the level of circumflex (Cx) artery (arrow). D = largest diameter.
short-axis view as well as the maximum (D2) diameters, and the maximum (D1) depth of LAA (the longest distance from LAA orifice to the tip of LAA and orifice) (Fig. 3). Statistical Analysis: Continuous variables are presented as percentages and means standard deviation. Categorical data are presented as absolute numbers and percentages. Continuous variables were compared using independent Student’s t-test; categorical variables were compared using chisquare test or Fisher’s exact test. A 2-sided Pvalue
