- Email: [email protected]
Assessment of Left Ventricular Mass in Hypertrophic Cardiomyopathy by Real-Time Three-Dimensional Echocardiography Using Single-Beat Capture Image
Assessment of Left Ventricular Mass in Hypertrophic Cardiomyopathy by Real-Time Three-Dimensional Echocardiography Using Single-Beat Capture Image Sung-A Chang, MD, Hyung-Kwan Kim, MD,* Sang-Chol Lee, MD,* Eun-Young Kim, RDCS, Seung-Hee Hahm, RDCS, Oh Min Kwon, RDCS, Seung Woo Park, MD, Yeon Hyeon Choe, MD, and Jae K. Oh, MD, Seoul, South Korea; Rochester, Minnesota
Background: Left ventricular (LV) mass is an important prognostic indicator in hypertrophic cardiomyopathy. Although LV mass can be easily calculated using conventional echocardiography, it is based on geometric assumptions and has inherent limitations in asymmetric left ventricles. Real-time three-dimensional echocardiographic (RT3DE) imaging with single-beat capture provides an opportunity for the accurate estimation of LV mass. The aim of this study was to validate this new technique for LV mass measurement in patients with hypertrophic cardiomyopathy. Methods: Sixty-nine patients with adequate two-dimensional (2D) and three-dimensional echocardiographic image quality underwent cardiac magnetic resonance (CMR) imaging and echocardiography on the same day. Real-time three-dimensional echocardiographic images were acquired using an Acuson SC2000 system, and CMR-determined LV mass was considered the reference standard. Left ventricular mass was derived using the formula of the American Society of Echocardiography (M-mode mass), the 2D-based truncated ellipsoid method (2D mass), and the RT3DE technique (RT3DE mass). Results: The mean time for RT3DE analysis was 5.85 6 1.81 min. Intraclass correlation analysis showed a close relationship between RT3DE and CMR LV mass (r = 0.86, P < .0001). However, LV mass by the M-mode or 2D technique showed a smaller intraclass correlation coefficient compared with CMR-determined mass (r = 0.48, P = .01, and r = 0.71, P < .001, respectively). Bland-Altman analysis showed reasonable limits of agreement between LV mass by RT3DE imaging and by CMR, with a smaller positive bias (19.5 g [9.1%]) compared with that by the M-mode and 2D methods (35.1 g [20.2%] and 30.6 g [17.6%], respectively). Conclusions: RT3DE measurement of LV mass using the single-beat capture technique is practical and more accurate than 2D or M-mode LV mass in patients with hypertrophic cardiomyopathy. (J Am Soc Echocardiogr 2013;26:436-42.) Keywords: Left ventricular mass, Three-dimensional echocardiography, Hypertrophic cardiomyopathy
From the Division of Cardiology, Department of Medicine (S.-A.C., S.-C.L., S.W.P., J.K.O.), and the Department of Radiology (Y.H.C.), Cardiovascular Imaging Center (S.-A.C., S.-C.L., E.-Y.K., S.-H.H., S.W.P., Y.H.C., J.K.O.), Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea; the Department of Internal Medicine, Cardiovascular Center, Seoul National University Hospital, Seoul, Korea (H.-K.K., O.M.K.); and the Division of Cardiovascular Diseases, Mayo Clinic College of Medicine, Rochester, Minnesota (J.K.O.). This study was supported by the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A070001). * Drs. Kim and Lee contributed equally. Reprint requests: Hyung-Kwan Kim, MD, PhD, Division of Cardiology, Department of Internal Medicine, Cardiovascular Center, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea (E-mail: [email protected]). 0894-7317/$36.00 Copyright 2013 by the American Society of Echocardiography. http://dx.doi.org/10.1016/j.echo.2012.12.015
436
Left ventricular (LV) mass is an important prognostic indicator of heart failure and sudden cardiac death in patients with hypertrophic cardiomyopathy (HCM).1,2 Although cardiac magnetic resonance (CMR) imaging is the most accurate noninvasive method for assessing myocardial mass in vivo,3 it is relatively expensive and time consuming and is limited in patients with claustrophobia or implanted pacemakers, limiting its routine use in the daily clinical practice. Transthoracic echocardiography is a widely used modality for assessing LV morphology and function in patients with HCM and can be easily performed for longitudinal assessment. Furthermore, it provides more accurate information on hemodynamic changes.4,5 LV mass determination using M-mode or two-dimensional (2D) echocardiography has been well validated in normal or hypertensive patients with symmetrically shaped left ventricles,6-8 but LV mass calculations using M-mode and 2D techniques are based on the assumption of symmetric LV geometry. Accordingly, it can be presumed that these conventional methods have inherent limitations that might produce erroneous results in patients with HCM. Three-dimensional (3D) assessments, which are free of geometric assumptions, might provide more accurate measures of LV mass in
Journal of the American Society of Echocardiography Volume 26 Number 4
asymmetric left ventricles than conventional 2D methods, but ASE = American Society of the effort required to acquire Echocardiography and analyze 3D echocardiographic data limits clinical utility. CMR = Cardiac magnetic However, the recent developresonance ment of real-time 3D echocarHCM = Hypertrophic diographic (RT3DE) imaging cardiomyopathy with single-beat capture has ICC = Intraclass correlation been shown to provide a straightcoefficient forward, highly accurate means of determining LV volume and LV = Left ventricular function.9 Additionally, this apRT3DE = Real-time threeproach provides a means for caldimensional culating LV mass that is free of echocardiographic geometric assumptions. We hypothesized that the ac3D = Three-dimensional curacy of LV mass calculation in 2D = Two-dimensional patients with HCM could be improved using RT3DE imaging with single-beat capture compared with conventional methods such as M-mode or 2D imaging. We also examined the feasibility and accuracy of this novel RT3DE technique for the determination of LV mass in patients with HCM, with CMR as the reference standard. Abbreviations
METHODS Study Population Consecutive patients with HCM in normal sinus rhythm at two tertiary referral hospitals (Samsung Medical Center and Seoul National University Hospital) with established diagnoses of HCM and scheduled for CMR and echocardiography on the same day were considered for this study. Patients with poor 2D echocardiographic windows, defined as those in whom an experienced sonographer could not sufficiently discriminate the endocardial and/or epicardial borders on 2D echocardiographic images, and patients with contraindications to CMR were excluded a priori. Ultimately, 71 patients were enrolled, and these patients constituted the study cohort. The study protocol was approved by the institutional review boards of the two participating hospitals. CMR Imaging All patients underwent CMR studies using a 1.5-Tscanner (Magnetom Avanto, syngo MR; Siemens Healthcare, Erlangen, Germany) during repeated breath holds. After localization, cine images for LV and right ventricular functional parameters were acquired using a steady-state free precession sequence (repetition time, 8–10 msec; echo time, 3–5 msec; flip angle, 20 ; in-plane resolution, 1.4 to 1.6 mm 2.2 to 2.5 mm; temporal resolution, 46 6 8 msec) with eight to 10 contiguous short-axis slices to cover the entire left and right ventricles, with slice thickness of 6 mm and 4-mm gaps.10 Image analysis was performed to determine LV mass from CMR images using commercial software (Argus version 4.02; Siemens Healthcare) by a single experienced observer blinded to all echocardiographic results. In each case, the end-diastolic frame with the largest LV cavity size was selected for LV mass measurement by retrospective image review. Endocardial and epicardial borders were manually traced in the selected image frame for the LV cavity volume and total LV volume (defined as the sum of LV cavity volume and LV myocardial volume) calculations. Papillary muscles and LV tra-
Chang et al 437
beculae were excluded from endocardium and included in LV cavity volume. At the base of the heart, slices were considered to be within the LV if the blood volume was surrounded by $50% of ventricular myocardium. Left ventricular myocardial volume was calculated by subtracting LV cavity volume from total LV volume. Finally, LV mass was calculated by multiplying LV myocardial volume by myocardial density (1.05 g/mL). Echocardiographic Image Acquisition After CMR acquisition, transthoracic echocardiography was performed using commercially available equipment (Acuson SC2000; Siemens Medical Solutions USA, Inc., Mountain View, CA) by a single experienced sonographer at each center, with subjects in the left lateral decubitus position. After each routine echocardiographic examination, 2D targeted M-mode echocardiographic images were acquired at the level of the mitral valve leaflet tips, in accordance with the American Society of Echocardiography (ASE) guidelines for chamber quantification.7 To calculate LV mass using 2D echocardiograms, short-axis view images at the mid ventricle and apical fourchamber view images were acquired. Great care was taken to avoid foreshortened acquisition of apical images. For LV mass quantification by RT3DE imaging, a special 3D image acquisition transducer (4Z1c) was used. This transducer has a matrix array with a maximum volume angle of 90 90 . Volume angles and frame rates were optimized for each patient to allow visualization of both endocardial and epicardial borders. Echocardiographic Analysis of LV Mass Measurement Echocardiographic images were transferred to a central laboratory and analyzed by two experienced observers (E.-Y.K. and S.-H.H), who independently analyzed M-mode, 2D, and 3D echocardiographic data without knowledge of CMR data. Analyses were performed using 2D and 3D software analysis packages supplied with the echocardiographic system. For LV mass measurement using the M-mode technique, we used the formula suggested by the ASE8: h i LV mass ðgÞ ¼ 0:80 1:04ðPWT þ LVIDd þ SWTÞ3 LVID3 þ 0:6 g;
where PWT is LV end-diastolic posterior wall thickness (millimeters), LVIDd is LVend-diastolic dimension (millimeters), and SWT is LVenddiastolic septal wall thickness (millimeters). For LV mass calculations by 2D echocardiography, we used the area-length method, as described in an ASE document on LV quantitation11; LV mass was calculated by subtracting endocardial volume from epicardial volume: LV mass ðgÞ ¼ 1:05 ½5=6ðAepi Lepi Þ ðAendo Lendo Þ :
For LV mass quantification by RT3DE imaging, analysis was performed online using an embedded program (Volume Cardiac Analysis Package – Volume Left Ventricular Analysis version 1.6) on the SC2000 system. In each case, a suitable end-diastolic frame depicting the maximally dilated LV cavity was chosen (Figure 1). After applying the automatic contouring algorithm, several manual corrections were made by visual assessment to adjust the endocardial border. After completing the endocardial tracing, the epicardial border was manually traced. On the basis of these tracings, the 3D analysis program in the platform calculated LV myocardial volume and provided a value for LV mass (grams). Time required for the analysis
438 Chang et al
Journal of the American Society of Echocardiography April 2013
Figure 1 Three-dimensional echocardiography and the analysis of LV mass. (A) An end-diastolic frame, containing a maximally dilated LV cavity, was selected by retrospective review. RT3DE images were reviewed in four planes (short-axis view and apical two-chamber, three-chamber, and four-chamber views), and brightness and contrast were optimized. (B) After the automatic contouring algorithm for endocardial border tracking was applied, several manual corrections were applied with visual assessment to adjust correct tracing for endocardial border. (C) After completion of endocardial tracing, the epicardial border was manually traced. was recorded for each case. Real-time three-dimensional echocardiographic image quality was graded as good (the endocardium and epicardium were entirely visualized) or suboptimal but interpretable (one segment of the endocardium or epicardium was not visualized, but tracing was possible by interpolation). For interobserver variability of LV mass assessment, echocardiographic images in 12 randomly selected patients were analyzed by two independent reviewers blinded to each other’s measurement. These measurements were also repeated by one reviewer blinded to the first measurement, at least 1 month apart from each measurement. Statistical Analysis Left ventricular mass data obtained using the different echocardiographic techniques and CMR are presented as mean 6 SD. Agreement is expressed using Bland-Altman plots with mean differences and 95% limits of agreements. Intraclass correlation coefficients (ICC) were used for comparisons among different measurement techniques. Intraobserver and interobserver variability was defined as the absolute difference between the corresponding repeated measurements expressed as a percentage of their mean. Variability values obtained for each parameter in each patient for each observer were averaged over the whole group of patients and are expressed as mean 6 SD. Statistical significance was accepted for P values < .05.
RESULTS Study Population Seventy-one consecutive patients were enrolled, but two patients were excluded from final analyses because of poor RT3DE image
Table 1 Characteristics of the study population (n = 69) Variable
Age (y) Men Type of HCM Septal Apical Septal plus apical Diffuse Others LV mass (g) ASE formula 2D ellipsoid method RT3DE imaging CMR LV ESV (mL) by CMR LV EDV (mL) by CMR LV EF (mL) by CMR
Value
58.2 6 10.9 58 (83%) 17 (24.6%) 32 (46.4%) 10 (14.5%) 3 (4.3%) 7 (10.1%) 212.8 6 61.5 148.3 6 49.1 159.4 6 45.8 178.9 6 64.4 41.2 6 16.4 138.8 6 29.9 70.3 6 8.7
EDV, End-diastolic volume; EF, ejection fraction; ESV, end-systolic volume. Data are expressed as mean 6 SD or as number (percentage).
quality. Characteristics of the study subjects are briefly summarized in Table 1. Apical hypertrophy was the most common HCM subtype, and most of the study subjects were men. Mean LV mass determined using the M-mode formula was greater than that measured by other modalities (CMR, 2D echocardiography, and RT3DE imaging; Table 1).
Chang et al 439
Journal of the American Society of Echocardiography Volume 26 Number 4
Figure 2 LV mass measured by M-mode and 2D echocardiography, RT3DE imaging, and CMR. (A) LV mass determined using the ASE formula was found to be moderately correlated with CMR-determined LV mass, though a large negative bias toward the reference method (CMR) was observed in the Bland-Altman plot. (B) LV mass determined using the 2D ellipsoid method showed better agreement with the CMR value than that obtained using the ASE formula, but a large positive bias and several results extreme outliers were observed. (C) LV mass determined using RT3DE imaging with single-beat capture showed good agreement with the CMR value and better limits of agreement (LOA) with less bias in the Bland-Altman plot than the other two methods. Feasibility of and Times Required for LV Mass Analysis by RT3DE Imaging When 2D echocardiographic image quality was satisfactory (n = 71), LV mass analysis by RT3DE imaging was possible in 97% of patients (n = 69). Image quality was suboptimal in 37 of these 69 patients (53.6%), and in most of these, border interpolation involved epicardial border tracing of apical segments in cases of apical HCM. The mean time required for RT3DE LV mass analysis was 5.85 61.81 min (range, 3–12.5 min), and the mean frame rate required for optimal image acquisition was 16.7 6 4.8 frames/sec.
LV Mass as Determined by M-Mode Echocardiography, 2D Echocardiography, and RT3DE Imaging Compared with CMR A summary regarding comparisons of LV mass data acquired using the different echocardiographic methods and CMR is provided in Figure 2. Intraclass correlation analysis showed a close association between RT3DE and CMR LV mass (r = 0.86, P < .0001), with a relatively narrow distribution along the reference line, highlighting a high concordance with LV mass by CMR (Figure 2C). Left ventricular masses determined by the M-mode and 2D techniques were less well correlated with CMR data (Figures 2A and 2B). Left ventricular mass by M-mode analysis displayed only a moderate correlation with LV mass by CMR analysis (r = 0.48, P < .001), with a relatively wide distribution and a large negative bias in the Bland-Altman plot (Figure 2A). In addition, although LV mass determined using the 2D ellipsoid method showed better agreement with CMR-determined
values in comparison with those by the M-mode method, the ICC for the 2D method (r = 0.82, P < .001) was lower than for the RT3DE method (r = 0.90, P < .001) (Table 2). Furthermore, in several cases, LV mass was considerably different from the CMR-determined value (Figure 2B). On the other hand, LV mass determined by RT3DE imaging displayed a substantially better correlation with CMRdetermined values (r = 0.86, P < .001), had a narrower distribution along the reference line, and showed little bias (Figure 2C). We also analyzed the ICC of each modality for LV mass measurement in relation to HCM type (nonapical vs apical; Table 3). Of note, RT3DE imaging was more beneficial in the nonapical type, compared with the apical type, in terms of LV mass measurement. Left ventricular mass assessed by the M-mode or 2D techniques showed lower ICCs for the nonapical type than for the apical type. Intraobserver and Interobserver Variability Intraobserver variability was 4.9 6 2.9%, 5.37 6 5.0%, and 5.0 6 5.2% for LV measurements by M-mode, 2D, and RT3DE imaging. Interobserver variability was 10.2 6 7.3%, 13.1 6 8.6%, and 6.3 6 4.5%, respectively.
DISCUSSION This is the first study to evaluate the accuracy and feasibility of LV mass measurements in patients with HCM using online analysis of RT3DE images with single-beat full-volume capture technology. Our comparison with conventional M-mode or 2D
440 Chang et al
Journal of the American Society of Echocardiography April 2013
Table 2 Summary of intermodality correlation and bias
ICC r P
M-mode vs CMR
2D vs CMR
RT3DE imaging vs CMR
0.66 <.001
0.82 <.001
0.90 <.001
echocardiographic analyses using CMR as a reference modality showed that RT3DE imaging more accurately determines LV mass in patients with HCM. The clinical feasibility and the higher accuracy of RT3DE LV mass measurements are of substantial clinical relevance, because LV mass is considered an important prognostic indicator in HCM. In addition, freedom of the assumption of symmetry is another important advantage of the RT3DE technique for LV mass measurement. This is of particular importance in patients with HCM, because its various phenotypic expressions make geometric assumptions implausible and thus make estimations based on these assumptions inherently inaccurate.
Limitations of Conventional Echocardiographic Methods for LV Mass Evaluation in Patients with HCM Left ventricular hypertrophy is an important risk factor of many cardiovascular diseases, and therefore, researchers have made continued efforts to devise a noninvasive, accurate means of assessing LV mass. However, previously described methods involve calculations based on geometric assumptions. For example, the formula advocated by the ASE assumes that the left ventricle is spherical, which clearly differs from its true shape. Nevertheless, a previous comparative study that used necropsy findings as a reference showed an excellent correlation (r = 0.90, P < .001)8 when it was applied to symmetrically shaped left ventricles. Another approach, the area-length method, supposes that the left ventricle is conically shaped,6 which seems a better assumption than the spherical shape, and not surprisingly, the data obtained using the area-length method were found to be better correlated with CMR-based LV mass data in comparison with those obtained using the M-mode method. However, the area-length method considers only the midwall plane of the left ventricle and thus does not take into account the presence of regional hypertrophy localized to basal and/or apical walls, frequently observed findings in patients with HCM. Left ventricular mass determination by CMR is also limited by partial volume artifacts and difficulties associated with the discrimination of right ventricular trabeculation and the septal wall. Nevertheless, it is a simple volumetric method that applies multiple stack analysis to images with high signal-to-noise ratios and acceptable spatial resolution, which better delineate the blood-endocardium border because of high signal difference between blood and myocardium. Currently, CMR is considered the most accurate in vivo technique for LV mass assessment and has been reported to be more accurate than echocardiographic methods in this respect.3,12 One small-scale comparative evaluation of LV mass as determined by echocardiography and CMR in patients with HCM showed a wide discrepancy between echocardiography and CMR but showed only small differences in normal subjects.13 Furthermore, the findings of this previous study suggest that the asymmetric shape of the left ventricle contributes to the inaccuracies of echocardiographic estimations of LV mass. Left ventricular mass estimation using 3D echocardiography adopts volumetric algorithms, which are more powerful for deter-
mining LV mass because they do not involve geometric assumptions. Furthermore, the CMR method for LV mass assessment is based on the summation of 2D image stacks, but there are gaps between imaging planes, and thus 3D echocardiography may be more accurate in vivo than CMR for determining LV myocardial mass. Feasibility, Image Quality, and Technical Pitfalls of RT3DE Imaging Multiple-beat RT3DE imaging has already been shown to significantly improve accuracy in LV mass quantification in comparison with 2D approaches, with CMR as a reference, irrespective of geometric assumptions or using manual adjustment for accurate delineation of endocardial and epicardial borders.14-19 These studies included about 20 patients and acquired RT3DE images with four wedged subvolumes and analyzed them offline. None of these studies included patients with HCM. Our study has some strengths, compared with the previous works. First, we tested a novel RT3DE technology in our ‘‘real-world’’ clinical practice. Our images were analyzed online using a program that has been already installed in echocardiographic equipment. We also provide data on analysis time, suggesting that RT3DE imaging with single-beat capture is a strong candidate for incorporation into daily clinical echocardiographic practice. Second, we showed the clinical utility of RT3DE imaging in a highly specified, unique HCM population. Because HCM is characterized by asymmetric hypertrophy of the left ventricle, geometric assumptions underlying LV mass quantification by 2D or M-mode methods do not seem to be appropriately applied, which is in clear contrast in patients with symmetric hypertrophy of the left ventricle, as in hypertensive LV hypertrophy. Previously published works involved small numbers of study patients and included a variety of clinical conditions. Hence, direct application of those results to the real-world clinical arena appears to be premature. Bicudo et al.20 reported LV mass quantification using RT3DE imaging in patients with HCM, but they used a multiple-beat capture 3D technique with offline analysis in a small number of patients, a significant drawback for clinical application. Our data showed, to some extent, inferior results compared with previous studies in LV mass quantification, which can be best explained by differences in study populations compared with those in the Western world. In Western countries, apical HCM is extremely rare, but it is common in Asian countries, which may also be true in the present study. Of interest, Table 3 highlights that LV mass quantification by RT3DE imaging is more beneficial in nonapical HCM than in apical HCM. The more asymmetric shape of the nonapical type may be a possible explanation. Another possible explanation may be poor delineation of the LV apex in apical HCM. Moreover, the epicardium of the hypertrophied LV apex usually bulges outward and presents some technical difficulties in accurate measurement. Furthermore, the current version of single-beat RT3DE imaging used in our study did not allow us to obtain images >90 in width, and thus the epicardial border of the hypertrophied apex can sometimes drop out or be vaguely addressed. In the present study, we showed better feasibility than that reported in a previous study by our group using RT3DE imaging with single-beat capture technology.9 In the previous study, we included many patients with dilated left ventricles or dilated apexes, which are not issues for most subjects with HCM, rendering the feasibility of this study much better than that of the previous work. Despite the high feasibility of RT3DE imaging in this study, the ICC for LV
Chang et al 441
Journal of the American Society of Echocardiography Volume 26 Number 4
Table 3 LV mass measurement according to HCM type: nonapical versus apical Nonapical (n = 37)
ICC (reference modality: CMR) r P
Apical (n = 32)
M-mode
2D
RT3DE imaging
M-mode
2D
RT3DE imaging
0.49 .03
0.73 <.001
0.91 <.001
0.81 <.001
0.87 <.001
0.85 <.001
mass measurement between RT3DE and CMR imaging was <0.9, slightly lower than the results from previous studies performed in symmetric left ventricles using multiple-beat capture RT3DE algorithms.21,22 In the present study, we recruited a large proportion of pure apical or apical-septal HCM subtypes (>50% of the study population). Echocardiographic images in apical HCM were more or less limited in completely demonstrating apical morphology compared with those of septal HCM, because the epicardial contour shows bulging in some cases; hence, epicardial tracing of the apex is not uncommonly challenging, and this hurdle might be a potential cause of underestimation of LV mass in this HCM subtype. Actually, about half of our study subjects required interpolation of the epicardial border during the tracing procedure. Although the mean time required for LV mass measurement by RT3DE imaging was about 5 min, it varied from 3 to 12 min, and the amount of time taken was determined primarily by how effectively the automatic contouring algorithm delineated the endocardial border and thus how much the time for manual correction of border delineation could be minimized. Another practical limitation is that the contemporary analysis program does not support the automatic contouring algorithm for epicardial border tracing but provides only a crude outline of the epicardial border. Future advances in automatic contouring techniques for epicardial border tracing are expected to shorten the time required for analysis, which facilitates rapid incorporation of this novel technique for LV measurement into daily echocardiographic practice.
Limitations The subjects enrolled in our study were all Korean. Thus, a relatively high proportion of apical HCM was included in the present study. This cohort makeup may have affected study results and may limit extrapolation of the present results to patients with the HCM type frequently observed in Western countries, because asymmetric septal hypertrophy is more prevalent in the West. However, as we demonstrated, RT3DE imaging was more beneficial for LV mass measurement in nonapical hypertrophy than apical hypertrophy. Real-time three-dimensional echocardiographic imaging with single-beat capture has potential advantages in terms of avoiding stitch artifacts, especially in patients with arrhythmias such as atrial fibrillation. However, in the present study, we included only patients in sinus rhythm, and thus our results may not be extended to patients with HCM with atrial fibrillation. Nevertheless, the single-beat capture approach likely holds promise, given its freedom from stitch artifacts, an inherent limitation of multiple-beat RT3DE imaging. This should be further investigated. Head-to-head comparison between the two approaches (i.e., single-beat capture vs multiple-beat capture) would be interesting in terms of LV mass measurements in patients with HCM with atrial fibrillation.
CONCLUSIONS Although conventional echocardiography provides a useful, noninvasive means for evaluating hemodynamics and ventricular function in patients with HCM, its accuracy in determining LV mass is limited. The present study shows that RT3DE imaging with single-beat capture is both more feasible and more accurate than 2D or M-mode echocardiography in terms of LV mass assessment, especially in nonapical HCM. Real-time three-dimensional echocardiographic measurements were more reproducible than 2D or M-mode echocardiography. Thus, we expect this technique to be incorporated into daily clinical practice in the near future.
REFERENCES 1. Spirito P, Bellone P, Harris KM, Bernabo P, Bruzzi P, Maron BJ. Magnitude of left ventricular hypertrophy and risk of sudden death in hypertrophic cardiomyopathy. N Engl J Med 2000;342:1778-85. 2. Elliott PM, Gimeno Blanes JR, Mahon NG, Poloniecki JD, McKenna WJ. Relation between severity of left-ventricular hypertrophy and prognosis in patients with hypertrophic cardiomyopathy. Lancet 2001;357:420-4. 3. Bottini PB, Carr AA, Prisant LM, Flickinger FW, Allison JD, Gottdiener JS. Magnetic resonance imaging compared to echocardiography to assess left ventricular mass in the hypertensive patient. Am J Hypertens 1995;8: 221-8. 4. Panza JA, Petrone RK, Fananapazir L, Maron BJ. Utility of continuous wave Doppler echocardiography in the noninvasive assessment of left ventricular outflow tract pressure gradient in patients with hypertrophic cardiomyopathy. J Am Coll Cardiol 1992;19:91-9. 5. Afonso LC, Bernal J, Bax JJ, Abraham TP. Echocardiography in hypertrophic cardiomyopathy: the role of conventional and emerging technologies. JACC Cardiovasc Imaging 2008;1:787-800. 6. Helak JW, Reichek N. Quantitation of human left ventricular mass and volume by two-dimensional echocardiography: in vitro anatomic validation. Circulation 1981;63:1398-407. 7. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18: 1440-63. 8. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450-8. 9. Chang SA, Lee SC, Kim EY, Hahm SH, Jang SY, Park SJ, et al. Feasibility of single-beat full-volume capture real-time three-dimensional echocardiography and auto-contouring algorithm for quantification of left ventricular volume: validation with cardiac magnetic resonance imaging. J Am Soc Echocardiogr 2011;24:853-9. 10. Natori S, Lai S, Finn JP, Gomes AS, Hundley WG, Jerosch-Herold M, et al. Cardiovascular function in multi-ethnic study of atherosclerosis: normal
442 Chang et al
11.
12.
13.
14.
15.
16.
values by age, sex, and ethnicity. AJR Am J Roentgenol 2006;186(suppl 2): S357-65. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, et al., American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. J Am Soc Echocardiogr 1989;2: 358-67. Katz J, Milliken MC, Stray-Gundersen J, Buja LM, Parkey RW, Mitchell JH, et al. Estimation of human myocardial mass with MR imaging. Radiology 1988;169:495-8. Allison JD, Flickinger FW, Wright JC, Falls DG III, Michael Prisant L, VonDohlen TW, et al. Measurement of left ventricular mass in hypertrophic cardiomyopathy using MRI: comparison with echocardiography. Magn Reson Imaging 1993;11:329-34. Mor-Avi V, Sugeng L, Weinert L, MacEneaney P, Caiani EG, Koch R, et al. Fast measurement of left ventricular mass with real-time three-dimensional echocardiography: comparison with magnetic resonance imaging. Circulation 2004;110:1814-8. Caiani EG, Corsi C, Sugeng L, MacEneaney P, Weinert L, Mor-Avi V, et al. Improved quantification of left ventricular mass based on endocardial and epicardial surface detection with real time three dimensional echocardiography. Heart 2006;92:213-9. Oe H, Hozumi T, Arai K, Matsumura Y, Negishi K, Sugioka K, et al. Comparison of accurate measurement of left ventricular mass in patients with hypertrophied hearts by real-time three-dimensional echocardiography versus magnetic resonance imaging. Am J Cardiol 2005;95:1263-7.
Journal of the American Society of Echocardiography April 2013
17. Qin JX, Jones M, Travaglini A, Song JM, Li J, White RD, et al. The accuracy of left ventricular mass determined by real-time three-dimensional echocardiography in chronic animal and clinical studies: a comparison with postmortem examination and magnetic resonance imaging. J Am Soc Echocardiogr 2005;18:1037-43. 18. van den Bosch AE, Robbers-Visser D, Krenning BJ, McGhie JS, Helbing WA, Meijboom FJ, et al. Comparison of real-time three-dimensional echocardiography to magnetic resonance imaging for assessment of left ventricular mass. Am J Cardiol 2006;97:113-7. 19. Qi X, Cogar B, Hsiung MC, Nanda NC, Miller AP, Yelamanchili P, et al. Live/real time three-dimensional transthoracic echocardiographic assessment of left ventricular volumes, ejection fraction, and mass compared with magnetic resonance imaging. Echocardiography 2007;24:166-73. 20. Bicudo LS, Tsutsui JM, Shiozaki A, Rochitte CE, Arteaga E, Mady C, et al. Value of real time three-dimensional echocardiography in patients with hypertrophic cardiomyopathy: comparison with two-dimensional echocardiography and magnetic resonance imaging. Echocardiography 2008;25:717-26. 21. Jenkins C, Bricknell K, Hanekom L, Marwick TH. Reproducibility and accuracy of echocardiographic measurements of left ventricular parameters using real-time three-dimensional echocardiography. J Am Coll Cardiol 2004;44:878-86. 22. Takeuchi M, Nishikage T, Mor-Avi V, Sugeng L, Weinert L, Nakai H, et al. Measurement of left ventricular mass by real-time three-dimensional echocardiography: validation against magnetic resonance and comparison with two-dimensional and m-mode measurements. J Am Soc Echocardiogr 2008;21:1001-5.
Background: Left ventricular (LV) mass is an important prognostic indicator in hypertrophic cardiomyopathy. Although LV mass can be easily calculated using conventional echocardiography, it is based on geometric assumptions and has inherent limitations in asymmetric left ventricles. Real-time three-dimensional echocardiographic (RT3DE) imaging with single-beat capture provides an opportunity for the accurate estimation of LV mass. The aim of this study was to validate this new technique for LV mass measurement in patients with hypertrophic cardiomyopathy. Methods: Sixty-nine patients with adequate two-dimensional (2D) and three-dimensional echocardiographic image quality underwent cardiac magnetic resonance (CMR) imaging and echocardiography on the same day. Real-time three-dimensional echocardiographic images were acquired using an Acuson SC2000 system, and CMR-determined LV mass was considered the reference standard. Left ventricular mass was derived using the formula of the American Society of Echocardiography (M-mode mass), the 2D-based truncated ellipsoid method (2D mass), and the RT3DE technique (RT3DE mass). Results: The mean time for RT3DE analysis was 5.85 6 1.81 min. Intraclass correlation analysis showed a close relationship between RT3DE and CMR LV mass (r = 0.86, P < .0001). However, LV mass by the M-mode or 2D technique showed a smaller intraclass correlation coefficient compared with CMR-determined mass (r = 0.48, P = .01, and r = 0.71, P < .001, respectively). Bland-Altman analysis showed reasonable limits of agreement between LV mass by RT3DE imaging and by CMR, with a smaller positive bias (19.5 g [9.1%]) compared with that by the M-mode and 2D methods (35.1 g [20.2%] and 30.6 g [17.6%], respectively). Conclusions: RT3DE measurement of LV mass using the single-beat capture technique is practical and more accurate than 2D or M-mode LV mass in patients with hypertrophic cardiomyopathy. (J Am Soc Echocardiogr 2013;26:436-42.) Keywords: Left ventricular mass, Three-dimensional echocardiography, Hypertrophic cardiomyopathy
From the Division of Cardiology, Department of Medicine (S.-A.C., S.-C.L., S.W.P., J.K.O.), and the Department of Radiology (Y.H.C.), Cardiovascular Imaging Center (S.-A.C., S.-C.L., E.-Y.K., S.-H.H., S.W.P., Y.H.C., J.K.O.), Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea; the Department of Internal Medicine, Cardiovascular Center, Seoul National University Hospital, Seoul, Korea (H.-K.K., O.M.K.); and the Division of Cardiovascular Diseases, Mayo Clinic College of Medicine, Rochester, Minnesota (J.K.O.). This study was supported by the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare & Family Affairs, Republic of Korea (A070001). * Drs. Kim and Lee contributed equally. Reprint requests: Hyung-Kwan Kim, MD, PhD, Division of Cardiology, Department of Internal Medicine, Cardiovascular Center, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, Seoul 110-744, Korea (E-mail: [email protected]). 0894-7317/$36.00 Copyright 2013 by the American Society of Echocardiography. http://dx.doi.org/10.1016/j.echo.2012.12.015
436
Left ventricular (LV) mass is an important prognostic indicator of heart failure and sudden cardiac death in patients with hypertrophic cardiomyopathy (HCM).1,2 Although cardiac magnetic resonance (CMR) imaging is the most accurate noninvasive method for assessing myocardial mass in vivo,3 it is relatively expensive and time consuming and is limited in patients with claustrophobia or implanted pacemakers, limiting its routine use in the daily clinical practice. Transthoracic echocardiography is a widely used modality for assessing LV morphology and function in patients with HCM and can be easily performed for longitudinal assessment. Furthermore, it provides more accurate information on hemodynamic changes.4,5 LV mass determination using M-mode or two-dimensional (2D) echocardiography has been well validated in normal or hypertensive patients with symmetrically shaped left ventricles,6-8 but LV mass calculations using M-mode and 2D techniques are based on the assumption of symmetric LV geometry. Accordingly, it can be presumed that these conventional methods have inherent limitations that might produce erroneous results in patients with HCM. Three-dimensional (3D) assessments, which are free of geometric assumptions, might provide more accurate measures of LV mass in
Journal of the American Society of Echocardiography Volume 26 Number 4
asymmetric left ventricles than conventional 2D methods, but ASE = American Society of the effort required to acquire Echocardiography and analyze 3D echocardiographic data limits clinical utility. CMR = Cardiac magnetic However, the recent developresonance ment of real-time 3D echocarHCM = Hypertrophic diographic (RT3DE) imaging cardiomyopathy with single-beat capture has ICC = Intraclass correlation been shown to provide a straightcoefficient forward, highly accurate means of determining LV volume and LV = Left ventricular function.9 Additionally, this apRT3DE = Real-time threeproach provides a means for caldimensional culating LV mass that is free of echocardiographic geometric assumptions. We hypothesized that the ac3D = Three-dimensional curacy of LV mass calculation in 2D = Two-dimensional patients with HCM could be improved using RT3DE imaging with single-beat capture compared with conventional methods such as M-mode or 2D imaging. We also examined the feasibility and accuracy of this novel RT3DE technique for the determination of LV mass in patients with HCM, with CMR as the reference standard. Abbreviations
METHODS Study Population Consecutive patients with HCM in normal sinus rhythm at two tertiary referral hospitals (Samsung Medical Center and Seoul National University Hospital) with established diagnoses of HCM and scheduled for CMR and echocardiography on the same day were considered for this study. Patients with poor 2D echocardiographic windows, defined as those in whom an experienced sonographer could not sufficiently discriminate the endocardial and/or epicardial borders on 2D echocardiographic images, and patients with contraindications to CMR were excluded a priori. Ultimately, 71 patients were enrolled, and these patients constituted the study cohort. The study protocol was approved by the institutional review boards of the two participating hospitals. CMR Imaging All patients underwent CMR studies using a 1.5-Tscanner (Magnetom Avanto, syngo MR; Siemens Healthcare, Erlangen, Germany) during repeated breath holds. After localization, cine images for LV and right ventricular functional parameters were acquired using a steady-state free precession sequence (repetition time, 8–10 msec; echo time, 3–5 msec; flip angle, 20 ; in-plane resolution, 1.4 to 1.6 mm 2.2 to 2.5 mm; temporal resolution, 46 6 8 msec) with eight to 10 contiguous short-axis slices to cover the entire left and right ventricles, with slice thickness of 6 mm and 4-mm gaps.10 Image analysis was performed to determine LV mass from CMR images using commercial software (Argus version 4.02; Siemens Healthcare) by a single experienced observer blinded to all echocardiographic results. In each case, the end-diastolic frame with the largest LV cavity size was selected for LV mass measurement by retrospective image review. Endocardial and epicardial borders were manually traced in the selected image frame for the LV cavity volume and total LV volume (defined as the sum of LV cavity volume and LV myocardial volume) calculations. Papillary muscles and LV tra-
Chang et al 437
beculae were excluded from endocardium and included in LV cavity volume. At the base of the heart, slices were considered to be within the LV if the blood volume was surrounded by $50% of ventricular myocardium. Left ventricular myocardial volume was calculated by subtracting LV cavity volume from total LV volume. Finally, LV mass was calculated by multiplying LV myocardial volume by myocardial density (1.05 g/mL). Echocardiographic Image Acquisition After CMR acquisition, transthoracic echocardiography was performed using commercially available equipment (Acuson SC2000; Siemens Medical Solutions USA, Inc., Mountain View, CA) by a single experienced sonographer at each center, with subjects in the left lateral decubitus position. After each routine echocardiographic examination, 2D targeted M-mode echocardiographic images were acquired at the level of the mitral valve leaflet tips, in accordance with the American Society of Echocardiography (ASE) guidelines for chamber quantification.7 To calculate LV mass using 2D echocardiograms, short-axis view images at the mid ventricle and apical fourchamber view images were acquired. Great care was taken to avoid foreshortened acquisition of apical images. For LV mass quantification by RT3DE imaging, a special 3D image acquisition transducer (4Z1c) was used. This transducer has a matrix array with a maximum volume angle of 90 90 . Volume angles and frame rates were optimized for each patient to allow visualization of both endocardial and epicardial borders. Echocardiographic Analysis of LV Mass Measurement Echocardiographic images were transferred to a central laboratory and analyzed by two experienced observers (E.-Y.K. and S.-H.H), who independently analyzed M-mode, 2D, and 3D echocardiographic data without knowledge of CMR data. Analyses were performed using 2D and 3D software analysis packages supplied with the echocardiographic system. For LV mass measurement using the M-mode technique, we used the formula suggested by the ASE8: h i LV mass ðgÞ ¼ 0:80 1:04ðPWT þ LVIDd þ SWTÞ3 LVID3 þ 0:6 g;
where PWT is LV end-diastolic posterior wall thickness (millimeters), LVIDd is LVend-diastolic dimension (millimeters), and SWT is LVenddiastolic septal wall thickness (millimeters). For LV mass calculations by 2D echocardiography, we used the area-length method, as described in an ASE document on LV quantitation11; LV mass was calculated by subtracting endocardial volume from epicardial volume: LV mass ðgÞ ¼ 1:05 ½5=6ðAepi Lepi Þ ðAendo Lendo Þ :
For LV mass quantification by RT3DE imaging, analysis was performed online using an embedded program (Volume Cardiac Analysis Package – Volume Left Ventricular Analysis version 1.6) on the SC2000 system. In each case, a suitable end-diastolic frame depicting the maximally dilated LV cavity was chosen (Figure 1). After applying the automatic contouring algorithm, several manual corrections were made by visual assessment to adjust the endocardial border. After completing the endocardial tracing, the epicardial border was manually traced. On the basis of these tracings, the 3D analysis program in the platform calculated LV myocardial volume and provided a value for LV mass (grams). Time required for the analysis
438 Chang et al
Journal of the American Society of Echocardiography April 2013
Figure 1 Three-dimensional echocardiography and the analysis of LV mass. (A) An end-diastolic frame, containing a maximally dilated LV cavity, was selected by retrospective review. RT3DE images were reviewed in four planes (short-axis view and apical two-chamber, three-chamber, and four-chamber views), and brightness and contrast were optimized. (B) After the automatic contouring algorithm for endocardial border tracking was applied, several manual corrections were applied with visual assessment to adjust correct tracing for endocardial border. (C) After completion of endocardial tracing, the epicardial border was manually traced. was recorded for each case. Real-time three-dimensional echocardiographic image quality was graded as good (the endocardium and epicardium were entirely visualized) or suboptimal but interpretable (one segment of the endocardium or epicardium was not visualized, but tracing was possible by interpolation). For interobserver variability of LV mass assessment, echocardiographic images in 12 randomly selected patients were analyzed by two independent reviewers blinded to each other’s measurement. These measurements were also repeated by one reviewer blinded to the first measurement, at least 1 month apart from each measurement. Statistical Analysis Left ventricular mass data obtained using the different echocardiographic techniques and CMR are presented as mean 6 SD. Agreement is expressed using Bland-Altman plots with mean differences and 95% limits of agreements. Intraclass correlation coefficients (ICC) were used for comparisons among different measurement techniques. Intraobserver and interobserver variability was defined as the absolute difference between the corresponding repeated measurements expressed as a percentage of their mean. Variability values obtained for each parameter in each patient for each observer were averaged over the whole group of patients and are expressed as mean 6 SD. Statistical significance was accepted for P values < .05.
RESULTS Study Population Seventy-one consecutive patients were enrolled, but two patients were excluded from final analyses because of poor RT3DE image
Table 1 Characteristics of the study population (n = 69) Variable
Age (y) Men Type of HCM Septal Apical Septal plus apical Diffuse Others LV mass (g) ASE formula 2D ellipsoid method RT3DE imaging CMR LV ESV (mL) by CMR LV EDV (mL) by CMR LV EF (mL) by CMR
Value
58.2 6 10.9 58 (83%) 17 (24.6%) 32 (46.4%) 10 (14.5%) 3 (4.3%) 7 (10.1%) 212.8 6 61.5 148.3 6 49.1 159.4 6 45.8 178.9 6 64.4 41.2 6 16.4 138.8 6 29.9 70.3 6 8.7
EDV, End-diastolic volume; EF, ejection fraction; ESV, end-systolic volume. Data are expressed as mean 6 SD or as number (percentage).
quality. Characteristics of the study subjects are briefly summarized in Table 1. Apical hypertrophy was the most common HCM subtype, and most of the study subjects were men. Mean LV mass determined using the M-mode formula was greater than that measured by other modalities (CMR, 2D echocardiography, and RT3DE imaging; Table 1).
Chang et al 439
Journal of the American Society of Echocardiography Volume 26 Number 4
Figure 2 LV mass measured by M-mode and 2D echocardiography, RT3DE imaging, and CMR. (A) LV mass determined using the ASE formula was found to be moderately correlated with CMR-determined LV mass, though a large negative bias toward the reference method (CMR) was observed in the Bland-Altman plot. (B) LV mass determined using the 2D ellipsoid method showed better agreement with the CMR value than that obtained using the ASE formula, but a large positive bias and several results extreme outliers were observed. (C) LV mass determined using RT3DE imaging with single-beat capture showed good agreement with the CMR value and better limits of agreement (LOA) with less bias in the Bland-Altman plot than the other two methods. Feasibility of and Times Required for LV Mass Analysis by RT3DE Imaging When 2D echocardiographic image quality was satisfactory (n = 71), LV mass analysis by RT3DE imaging was possible in 97% of patients (n = 69). Image quality was suboptimal in 37 of these 69 patients (53.6%), and in most of these, border interpolation involved epicardial border tracing of apical segments in cases of apical HCM. The mean time required for RT3DE LV mass analysis was 5.85 61.81 min (range, 3–12.5 min), and the mean frame rate required for optimal image acquisition was 16.7 6 4.8 frames/sec.
LV Mass as Determined by M-Mode Echocardiography, 2D Echocardiography, and RT3DE Imaging Compared with CMR A summary regarding comparisons of LV mass data acquired using the different echocardiographic methods and CMR is provided in Figure 2. Intraclass correlation analysis showed a close association between RT3DE and CMR LV mass (r = 0.86, P < .0001), with a relatively narrow distribution along the reference line, highlighting a high concordance with LV mass by CMR (Figure 2C). Left ventricular masses determined by the M-mode and 2D techniques were less well correlated with CMR data (Figures 2A and 2B). Left ventricular mass by M-mode analysis displayed only a moderate correlation with LV mass by CMR analysis (r = 0.48, P < .001), with a relatively wide distribution and a large negative bias in the Bland-Altman plot (Figure 2A). In addition, although LV mass determined using the 2D ellipsoid method showed better agreement with CMR-determined
values in comparison with those by the M-mode method, the ICC for the 2D method (r = 0.82, P < .001) was lower than for the RT3DE method (r = 0.90, P < .001) (Table 2). Furthermore, in several cases, LV mass was considerably different from the CMR-determined value (Figure 2B). On the other hand, LV mass determined by RT3DE imaging displayed a substantially better correlation with CMRdetermined values (r = 0.86, P < .001), had a narrower distribution along the reference line, and showed little bias (Figure 2C). We also analyzed the ICC of each modality for LV mass measurement in relation to HCM type (nonapical vs apical; Table 3). Of note, RT3DE imaging was more beneficial in the nonapical type, compared with the apical type, in terms of LV mass measurement. Left ventricular mass assessed by the M-mode or 2D techniques showed lower ICCs for the nonapical type than for the apical type. Intraobserver and Interobserver Variability Intraobserver variability was 4.9 6 2.9%, 5.37 6 5.0%, and 5.0 6 5.2% for LV measurements by M-mode, 2D, and RT3DE imaging. Interobserver variability was 10.2 6 7.3%, 13.1 6 8.6%, and 6.3 6 4.5%, respectively.
DISCUSSION This is the first study to evaluate the accuracy and feasibility of LV mass measurements in patients with HCM using online analysis of RT3DE images with single-beat full-volume capture technology. Our comparison with conventional M-mode or 2D
440 Chang et al
Journal of the American Society of Echocardiography April 2013
Table 2 Summary of intermodality correlation and bias
ICC r P
M-mode vs CMR
2D vs CMR
RT3DE imaging vs CMR
0.66 <.001
0.82 <.001
0.90 <.001
echocardiographic analyses using CMR as a reference modality showed that RT3DE imaging more accurately determines LV mass in patients with HCM. The clinical feasibility and the higher accuracy of RT3DE LV mass measurements are of substantial clinical relevance, because LV mass is considered an important prognostic indicator in HCM. In addition, freedom of the assumption of symmetry is another important advantage of the RT3DE technique for LV mass measurement. This is of particular importance in patients with HCM, because its various phenotypic expressions make geometric assumptions implausible and thus make estimations based on these assumptions inherently inaccurate.
Limitations of Conventional Echocardiographic Methods for LV Mass Evaluation in Patients with HCM Left ventricular hypertrophy is an important risk factor of many cardiovascular diseases, and therefore, researchers have made continued efforts to devise a noninvasive, accurate means of assessing LV mass. However, previously described methods involve calculations based on geometric assumptions. For example, the formula advocated by the ASE assumes that the left ventricle is spherical, which clearly differs from its true shape. Nevertheless, a previous comparative study that used necropsy findings as a reference showed an excellent correlation (r = 0.90, P < .001)8 when it was applied to symmetrically shaped left ventricles. Another approach, the area-length method, supposes that the left ventricle is conically shaped,6 which seems a better assumption than the spherical shape, and not surprisingly, the data obtained using the area-length method were found to be better correlated with CMR-based LV mass data in comparison with those obtained using the M-mode method. However, the area-length method considers only the midwall plane of the left ventricle and thus does not take into account the presence of regional hypertrophy localized to basal and/or apical walls, frequently observed findings in patients with HCM. Left ventricular mass determination by CMR is also limited by partial volume artifacts and difficulties associated with the discrimination of right ventricular trabeculation and the septal wall. Nevertheless, it is a simple volumetric method that applies multiple stack analysis to images with high signal-to-noise ratios and acceptable spatial resolution, which better delineate the blood-endocardium border because of high signal difference between blood and myocardium. Currently, CMR is considered the most accurate in vivo technique for LV mass assessment and has been reported to be more accurate than echocardiographic methods in this respect.3,12 One small-scale comparative evaluation of LV mass as determined by echocardiography and CMR in patients with HCM showed a wide discrepancy between echocardiography and CMR but showed only small differences in normal subjects.13 Furthermore, the findings of this previous study suggest that the asymmetric shape of the left ventricle contributes to the inaccuracies of echocardiographic estimations of LV mass. Left ventricular mass estimation using 3D echocardiography adopts volumetric algorithms, which are more powerful for deter-
mining LV mass because they do not involve geometric assumptions. Furthermore, the CMR method for LV mass assessment is based on the summation of 2D image stacks, but there are gaps between imaging planes, and thus 3D echocardiography may be more accurate in vivo than CMR for determining LV myocardial mass. Feasibility, Image Quality, and Technical Pitfalls of RT3DE Imaging Multiple-beat RT3DE imaging has already been shown to significantly improve accuracy in LV mass quantification in comparison with 2D approaches, with CMR as a reference, irrespective of geometric assumptions or using manual adjustment for accurate delineation of endocardial and epicardial borders.14-19 These studies included about 20 patients and acquired RT3DE images with four wedged subvolumes and analyzed them offline. None of these studies included patients with HCM. Our study has some strengths, compared with the previous works. First, we tested a novel RT3DE technology in our ‘‘real-world’’ clinical practice. Our images were analyzed online using a program that has been already installed in echocardiographic equipment. We also provide data on analysis time, suggesting that RT3DE imaging with single-beat capture is a strong candidate for incorporation into daily clinical echocardiographic practice. Second, we showed the clinical utility of RT3DE imaging in a highly specified, unique HCM population. Because HCM is characterized by asymmetric hypertrophy of the left ventricle, geometric assumptions underlying LV mass quantification by 2D or M-mode methods do not seem to be appropriately applied, which is in clear contrast in patients with symmetric hypertrophy of the left ventricle, as in hypertensive LV hypertrophy. Previously published works involved small numbers of study patients and included a variety of clinical conditions. Hence, direct application of those results to the real-world clinical arena appears to be premature. Bicudo et al.20 reported LV mass quantification using RT3DE imaging in patients with HCM, but they used a multiple-beat capture 3D technique with offline analysis in a small number of patients, a significant drawback for clinical application. Our data showed, to some extent, inferior results compared with previous studies in LV mass quantification, which can be best explained by differences in study populations compared with those in the Western world. In Western countries, apical HCM is extremely rare, but it is common in Asian countries, which may also be true in the present study. Of interest, Table 3 highlights that LV mass quantification by RT3DE imaging is more beneficial in nonapical HCM than in apical HCM. The more asymmetric shape of the nonapical type may be a possible explanation. Another possible explanation may be poor delineation of the LV apex in apical HCM. Moreover, the epicardium of the hypertrophied LV apex usually bulges outward and presents some technical difficulties in accurate measurement. Furthermore, the current version of single-beat RT3DE imaging used in our study did not allow us to obtain images >90 in width, and thus the epicardial border of the hypertrophied apex can sometimes drop out or be vaguely addressed. In the present study, we showed better feasibility than that reported in a previous study by our group using RT3DE imaging with single-beat capture technology.9 In the previous study, we included many patients with dilated left ventricles or dilated apexes, which are not issues for most subjects with HCM, rendering the feasibility of this study much better than that of the previous work. Despite the high feasibility of RT3DE imaging in this study, the ICC for LV
Chang et al 441
Journal of the American Society of Echocardiography Volume 26 Number 4
Table 3 LV mass measurement according to HCM type: nonapical versus apical Nonapical (n = 37)
ICC (reference modality: CMR) r P
Apical (n = 32)
M-mode
2D
RT3DE imaging
M-mode
2D
RT3DE imaging
0.49 .03
0.73 <.001
0.91 <.001
0.81 <.001
0.87 <.001
0.85 <.001
mass measurement between RT3DE and CMR imaging was <0.9, slightly lower than the results from previous studies performed in symmetric left ventricles using multiple-beat capture RT3DE algorithms.21,22 In the present study, we recruited a large proportion of pure apical or apical-septal HCM subtypes (>50% of the study population). Echocardiographic images in apical HCM were more or less limited in completely demonstrating apical morphology compared with those of septal HCM, because the epicardial contour shows bulging in some cases; hence, epicardial tracing of the apex is not uncommonly challenging, and this hurdle might be a potential cause of underestimation of LV mass in this HCM subtype. Actually, about half of our study subjects required interpolation of the epicardial border during the tracing procedure. Although the mean time required for LV mass measurement by RT3DE imaging was about 5 min, it varied from 3 to 12 min, and the amount of time taken was determined primarily by how effectively the automatic contouring algorithm delineated the endocardial border and thus how much the time for manual correction of border delineation could be minimized. Another practical limitation is that the contemporary analysis program does not support the automatic contouring algorithm for epicardial border tracing but provides only a crude outline of the epicardial border. Future advances in automatic contouring techniques for epicardial border tracing are expected to shorten the time required for analysis, which facilitates rapid incorporation of this novel technique for LV measurement into daily echocardiographic practice.
Limitations The subjects enrolled in our study were all Korean. Thus, a relatively high proportion of apical HCM was included in the present study. This cohort makeup may have affected study results and may limit extrapolation of the present results to patients with the HCM type frequently observed in Western countries, because asymmetric septal hypertrophy is more prevalent in the West. However, as we demonstrated, RT3DE imaging was more beneficial for LV mass measurement in nonapical hypertrophy than apical hypertrophy. Real-time three-dimensional echocardiographic imaging with single-beat capture has potential advantages in terms of avoiding stitch artifacts, especially in patients with arrhythmias such as atrial fibrillation. However, in the present study, we included only patients in sinus rhythm, and thus our results may not be extended to patients with HCM with atrial fibrillation. Nevertheless, the single-beat capture approach likely holds promise, given its freedom from stitch artifacts, an inherent limitation of multiple-beat RT3DE imaging. This should be further investigated. Head-to-head comparison between the two approaches (i.e., single-beat capture vs multiple-beat capture) would be interesting in terms of LV mass measurements in patients with HCM with atrial fibrillation.
CONCLUSIONS Although conventional echocardiography provides a useful, noninvasive means for evaluating hemodynamics and ventricular function in patients with HCM, its accuracy in determining LV mass is limited. The present study shows that RT3DE imaging with single-beat capture is both more feasible and more accurate than 2D or M-mode echocardiography in terms of LV mass assessment, especially in nonapical HCM. Real-time three-dimensional echocardiographic measurements were more reproducible than 2D or M-mode echocardiography. Thus, we expect this technique to be incorporated into daily clinical practice in the near future.
REFERENCES 1. Spirito P, Bellone P, Harris KM, Bernabo P, Bruzzi P, Maron BJ. Magnitude of left ventricular hypertrophy and risk of sudden death in hypertrophic cardiomyopathy. N Engl J Med 2000;342:1778-85. 2. Elliott PM, Gimeno Blanes JR, Mahon NG, Poloniecki JD, McKenna WJ. Relation between severity of left-ventricular hypertrophy and prognosis in patients with hypertrophic cardiomyopathy. Lancet 2001;357:420-4. 3. Bottini PB, Carr AA, Prisant LM, Flickinger FW, Allison JD, Gottdiener JS. Magnetic resonance imaging compared to echocardiography to assess left ventricular mass in the hypertensive patient. Am J Hypertens 1995;8: 221-8. 4. Panza JA, Petrone RK, Fananapazir L, Maron BJ. Utility of continuous wave Doppler echocardiography in the noninvasive assessment of left ventricular outflow tract pressure gradient in patients with hypertrophic cardiomyopathy. J Am Coll Cardiol 1992;19:91-9. 5. Afonso LC, Bernal J, Bax JJ, Abraham TP. Echocardiography in hypertrophic cardiomyopathy: the role of conventional and emerging technologies. JACC Cardiovasc Imaging 2008;1:787-800. 6. Helak JW, Reichek N. Quantitation of human left ventricular mass and volume by two-dimensional echocardiography: in vitro anatomic validation. Circulation 1981;63:1398-407. 7. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18: 1440-63. 8. Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450-8. 9. Chang SA, Lee SC, Kim EY, Hahm SH, Jang SY, Park SJ, et al. Feasibility of single-beat full-volume capture real-time three-dimensional echocardiography and auto-contouring algorithm for quantification of left ventricular volume: validation with cardiac magnetic resonance imaging. J Am Soc Echocardiogr 2011;24:853-9. 10. Natori S, Lai S, Finn JP, Gomes AS, Hundley WG, Jerosch-Herold M, et al. Cardiovascular function in multi-ethnic study of atherosclerosis: normal
442 Chang et al
11.
12.
13.
14.
15.
16.
values by age, sex, and ethnicity. AJR Am J Roentgenol 2006;186(suppl 2): S357-65. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, et al., American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. J Am Soc Echocardiogr 1989;2: 358-67. Katz J, Milliken MC, Stray-Gundersen J, Buja LM, Parkey RW, Mitchell JH, et al. Estimation of human myocardial mass with MR imaging. Radiology 1988;169:495-8. Allison JD, Flickinger FW, Wright JC, Falls DG III, Michael Prisant L, VonDohlen TW, et al. Measurement of left ventricular mass in hypertrophic cardiomyopathy using MRI: comparison with echocardiography. Magn Reson Imaging 1993;11:329-34. Mor-Avi V, Sugeng L, Weinert L, MacEneaney P, Caiani EG, Koch R, et al. Fast measurement of left ventricular mass with real-time three-dimensional echocardiography: comparison with magnetic resonance imaging. Circulation 2004;110:1814-8. Caiani EG, Corsi C, Sugeng L, MacEneaney P, Weinert L, Mor-Avi V, et al. Improved quantification of left ventricular mass based on endocardial and epicardial surface detection with real time three dimensional echocardiography. Heart 2006;92:213-9. Oe H, Hozumi T, Arai K, Matsumura Y, Negishi K, Sugioka K, et al. Comparison of accurate measurement of left ventricular mass in patients with hypertrophied hearts by real-time three-dimensional echocardiography versus magnetic resonance imaging. Am J Cardiol 2005;95:1263-7.
Journal of the American Society of Echocardiography April 2013
17. Qin JX, Jones M, Travaglini A, Song JM, Li J, White RD, et al. The accuracy of left ventricular mass determined by real-time three-dimensional echocardiography in chronic animal and clinical studies: a comparison with postmortem examination and magnetic resonance imaging. J Am Soc Echocardiogr 2005;18:1037-43. 18. van den Bosch AE, Robbers-Visser D, Krenning BJ, McGhie JS, Helbing WA, Meijboom FJ, et al. Comparison of real-time three-dimensional echocardiography to magnetic resonance imaging for assessment of left ventricular mass. Am J Cardiol 2006;97:113-7. 19. Qi X, Cogar B, Hsiung MC, Nanda NC, Miller AP, Yelamanchili P, et al. Live/real time three-dimensional transthoracic echocardiographic assessment of left ventricular volumes, ejection fraction, and mass compared with magnetic resonance imaging. Echocardiography 2007;24:166-73. 20. Bicudo LS, Tsutsui JM, Shiozaki A, Rochitte CE, Arteaga E, Mady C, et al. Value of real time three-dimensional echocardiography in patients with hypertrophic cardiomyopathy: comparison with two-dimensional echocardiography and magnetic resonance imaging. Echocardiography 2008;25:717-26. 21. Jenkins C, Bricknell K, Hanekom L, Marwick TH. Reproducibility and accuracy of echocardiographic measurements of left ventricular parameters using real-time three-dimensional echocardiography. J Am Coll Cardiol 2004;44:878-86. 22. Takeuchi M, Nishikage T, Mor-Avi V, Sugeng L, Weinert L, Nakai H, et al. Measurement of left ventricular mass by real-time three-dimensional echocardiography: validation against magnetic resonance and comparison with two-dimensional and m-mode measurements. J Am Soc Echocardiogr 2008;21:1001-5.