Accuracy and reproducibility of left ventricular mass measurement by subcostal M-mode echocardiography in hypertensive patients and professional bicyclists

In some patients, left ventricular (LV) mass cannot be evaluated by M-mode echocardiography because the parasternal long-axis view is not available. The aim of this study was to determine whether the subcostal view obtained by N-mode echocardiography under 2-dimensional guidance allows accurate and reproducible LV mass determination. Using the cube formula, LV mass was calculated, from parasternal and subcostal views in 96 subjects: 73 hypertensives and 23 professional bicyclists, covering a wide range of LV dimensions. M-mode tracings were read by 2 experienced echocardiographers and the interobserver variability was evaluated. With use of the subcostal view, the interobserver reproducibility, expressed as observer 1-observer 2, was excellent: -0.3 [+ or -] 1.3 mm for LV diastolic diameter, -0.1 [+ or -] 1.0 mm for ventricular diastolic septal thickness, 0.2 [+ or -] 0.6 mm for diastolic free wall thickness and 0.03 [+ or -] 16.7 g for LV mass. In 96% of cages, the difference in LV mass between the 2 observers did not exceed 30 g, With use of the parasternal and subcostal approaches, LV mass was not statistically different (202.6 [+ or -] 2.2 g and 206.5 [+ or -] 2.0 g, respectively) and the difference was Left ventricular (LV) hypertrophy has a recognized prognostic value in patients with systemic hypertension, and is an independent predictor of increased morbidity and mortality.[1-5] LV hypertrophy can be easily characterized and quantified by M-mode echocardiography under 2-dimensional guidance using a left parasternal view of the left ventricle.[6,7] LV mass determination, derived from an elliptical model of the left ventricle incorporating the cavity diameter and wall thickness, is reliable and reproducible,[8-10] providing that certain patients are excluded.[11] Patients with obesity, respiratory insufficiency or unaligned septum offer a poor quality parasternal view. hi such cases, the subcostal view may be useful since the M-mode ultrasonic beam can also be aimed perpendicularly to the long axis of the left ventricle.[12] Although some studies have compared LV dimensions using both approaches,[12-15] no study comparing the evaluation of LV mass by these 2 methods has been published. Therefore, the aim of the present study was to assess the accuracy and reproducibility of LV dimension and mass measurements obtained from a subcostal view compared with those obtained from the parasternal view. METHODS Enchocardiography: All examinations were performed using a Hewlett Packard Sonos 100 machine with a 2.5 MHz transducer. M-mode tracings of the left ventricle using parasternal and subcostal views were obtained from each patient during the same examination. We used the recommendations of Devereux regarding M-mode echocardiographic technique, using the parasternal window for measurement of the left ventricle.[11] For the subcostal view, special attention was paid to the following procedures: (1) accurate 2-dimensional imaging of LV long axis; this was achieved with patients in the left lateral position in the majority of cases, or in the supine position; (2) directing the ultrasonic M-mode beam perpendicularly to the walls, just below the tip of the mitral leaflet (using the same 'operator dependent' criteria as for the parasternal window); the M-mode echocardiographic recording was obtained during apnea at the moment in the respiratory cycle giving the optimal echo beam position (most often found to be at half to full inspiration); and (3) the maintenance of apnea, allowing the recording of [greater than or equal to] 3 cardiac cycles. Both the examination procedure and image interpretation were performed by 2 fully trained echocardiographers, who had previously achieved good interobserver reproducibility with this technique before the present study. Figure 1 shows the LV wall segments explored by both parasternal and subcostal M-mode examination. The following ventricular parameters were recorded on the parasternal and subcostal M-mode echocardiograms: LV diastolic and systolic internal dimensions, ventricular septal thickness, and free wall thickness. Inclusion design: Subjects were selected from 2 different populations: (1) hypertensive patients referred to our outpatient hypertension clinic, and (2) normotensive professional bicyclists participating in the 1991 Tour de France. We used the following inclusion criteria based on echographic data from a complete preliminary scan: (1) availability of an accurate M-mode echocardiogram of the left ventricle with both parasternal and subcostal views; (2) absence of LV distortion; and (3) absence of excessive variability in wall thickness (e.g., localized ventricular septal hypertrophy). Study design: No M-mode scans were interpreted until the completion of the study. Subjects' identification labels were masked and each echocardiogram interpreted independently. Each parameter was measured using the conventions of the American Society of Echocardiography[11] and averaged over 3 cardiac cycles by each of the 2 observers. Tracings were then unmasked, and the results of the 2 observers compared. The following rules were observed before any data analysis: (1) When the difference between the 2 observers' readings was Statistical analysis: Results are shown as mean [+ or -] 1 SD. Between-group comparisons were performed using the unpaired Student's t test. A p value RESULTS Study subjects: Table I summarizes the main characteristics of the 96 subjects included in this study. Only 4 subjects were excluded before data analysis, all due to poor quality paper tracings despite a satisfactory screen image during the on-line examination. [TABULAR DATA I OMITTED] Reproducibility of subcostal readings: The values of the different parameters measured by each observer are listed in Table II. For ventricular septal readings, significant discrepancies (difference >1 mm) were observed in 11 cases, the difference exceeding 2 mm in 5 cases (5.2%). For free wall thickness, the difference never exceeded 2 mm. For LV diastolic internal diameter, 3 readings differed by >2 mm, and 2 by >3 mm. The between-observer difference in LV mass did not exceed 30 g in 96% of cases, the greatest discrepancy being 65 g (Figure 2). LV mass difference did not depend on LV mass value. [TABULAR DATA II OMITTED] Comparison of left ventricular dimensions using parasternal and subcostal views: The main results are listed in Table III. Of 96 cases, the difference between parasternal and subcostal views for ventricular septal thickness was 2 mm in 12. For free wall thickness, the difference was >1 mm in 33 cases and >2 mm in 4 (4.2%). For LV diastolic internal diameter measurements, the difference was >2 mm, >3 mm or >5 mm in 58, 41 and 16%, respectively. The differences in the latter parameter ranged from -10 to +8 mm, with no over- or underestimation by the subcostal measurements compared with parasternal measurements (Figure 3). LV mass measurements exhibited a difference [TABULAR DATA III OMITTED] DISCUSSION Concerning the application of the same mathematic model to calculate mass using parastemal and subcostal view: LV mass can be accurately estimated by M-mode echocardiography, using the cube formula based on a prolate ellipsoid model (Figure 1) of a regularly configurated left ventricle (major to minor axis ratio approximately 2 to 1).[8,9] Myocardial volume is calculated as the difference between the external (external ellipsoid) and the internal (internal ellipsoid) volume of the left ventricle. LV volume may be approximated using only the minor axis measurement. Determination of the minor axis of the internal (LV diastolic internal diameter) and the external (LV diastolic internal diameter + wall segments) ellipsoids is derived from an M-mode echocardiogram of the left ventricle using the left parasternal window. The base of the ellipsoid is circular, which, in theory, allows an infinite number of identical measurements passing through the center. However, the number of echocardiographic views is limited and only the subcostal view allows the measurement of another minor axis of the left ventricle more or less perpendicular to that obtained using the parasternal view (Figure 1). Therefore, the subcostal view should allow the calculation of LV mass in nondeformed hearts. Reproducibility of readings: A measurement difference in reading of 1 mm for walls and 2 mm for the diameter is usually technically acceptable.[16,17] However, this can lead to variations in LV mass of up to 30 g for the same subject. In our clinic, we had previously evaluated the reproducibility of parasternal M-mode tracing interpretations in our daily practice. The mean between-observer difference in LV mass was 2 [+ or -] 8 g, a value comparable to that found in the present study using subcostal M-mode tracings. Finally, at the group level, the mean difference in LV mass between subcostal and parasternal views is not clinically relevant ( Comparison of the results obtained by each approach: Chang et al[12] and Starling et al[15] found comparable LV diastolic internal diameter values with both parasternal and subcostal views in ideal conditions. Bruntz et al[13] reported an overestimation using the parasternal approach (mean 53.4 vs 52.4 mm), whereas Rein et al[14] reported an overestimation using the subcostal view (51.4 [+ or -] 6.4 mm vs 50.0 [+ or -] 5.2 mm). The difference was statistically significant in Bruntz's study only; in none of these studies was the difference clinically relevant. Our results in the nondilated (hypertensive subjects) and in slightly dilated (bicyclists) left ventricle are in agreement with published data. To our knowledge, there is no reported data comparing LV mass calculation using parasternal and subcostal views. The present study shows similar results with both methods. This is corroborated by the fact that our SDs for each view are of the same order of magnitude as those obtained during reproducibility testing of repeated LV mass measurements using the parasternal view.[19] A case by case study shows a between-method difference of Study limitations: Only the reproducibility of readings was tested. We have no data on the inter- and intraexamination reproducibility of LV dimensions. In normal subjects, Brenner and Waugh[20] found a 2.9 mm inspiratory decrease of LV diastolic dimension; in this study, the phase of the respiratory cycle used for M-mode LV imaging was generally different for the 2 views but usually constant for each--mild to full expiration for the parasternal view and mild to full inspiration for the subcostal view. Clinical implications: The parasternal view remains the reference method for LV mass determination because it has been both anatomically validated and shown to be more reproducible than the subcostal view. However, in population studies, an M-mode parasternal view of the left ventricle is not achievable in 10 to 25% of cases,[11,15] especially in elderly and obese patients. Therefore, reliable LV assessment using another method is required. In this study concerning selected patients, the subcostal approach was a valuable alternative. In individual patients, this method allows the determination of LV mass for clinical purposes. In a research setting, recourse to the subcostal view may prevent patient exclusion because of a poor parasternal window. A prospective study in an unselected population would be justified to estimate the proportion of subjects in whom subcostal LV mass measurement can be performed accurately when the parasternal view is not reliable. Acknowledgment: We warmly thank Anne and Nick Berendt for their contribution in preparing the manuscript. [1.] Casale PN, Devereux RB, Milner M, Zullo G, Harshfield GA, Pickering TG, Laragh JH. Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann Intern Med 1986; 105:173-178. [2.] Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Left ventricular mass and incidence of coronary heart disease in an elderly cohort. Ann Intern Med 1989;110:101-107. [3.] Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990;322:1561-1566, [4.] Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med 1991;114:345-352. [5.] Ghali JK, Liao Y, Simmons B, Castaner A, Cao G, Cooper RS. The prognostic role of left ventricular hypertrophy in patients with or without coronary artery disease. Ann Intern Med 1992;117:831-836. [6.] Pearson AC, Gudipati CV, Nagelhout D, Sear J, Cohen JD, Labovitz AJ. Echocardiographic evaluation of cardiac structure and function in elderly subjects with isolated systolic hypertension. J Am Coll Cardiol 1991;17:422-430. [7.] Geva B, Elkayam U, Frishman WH. Determination of left ventricular wall thickening in patients with chronic systemic hypertension. Chest 1979;76:557-562. [8.] Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of left ventricular hypertrophy: comparison to necropsy findings. Am J Cardiol 1986;57:450-458. [9.] Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method. Circulation 1977;55:613-618. [10.] Collins HW, Kronenberg MW, Byrd BF. Reproducibility of left ventricular mass measurements by two-dimensional and M-mode echocardiography. J Am Coll Cardiol 1989;14:672-676. [11.] Devereux RB, Liebson PR, Horan MJ. Recommendations concerning the use of echocardiography in hypertension and general population research. Hypertension 1987;9(suppl II):II-97-11-104. [12.] Chang S, Feigenbaum H, Dillon J. Subxiphoid echocardiography. Chest 1975; 68:233-235. [13.] Bruntz JF, Chivoret G, Bourdon JL, Ethevenot G, Aliot E, Gilgenkrantz JM. Comparaison des principaux parametres echocardiographiques (mode TM) obtenus par voie parasternale et sous-costale. Arch Mal Coeur 1982;10:1159-1166. [14.] Rein AJJ, Azancot I, N'Gyen AVC, Adda JL, Georgiopoulos G, Piekarski A, Slama R. Subcostal M-mode computerised echocardiography, an alternative or complementary approach to parasternal echocardiography. Br Heart J 1983;50: 21-26. [15.] Starling MR, Crawford MH, O'Rourke RA, Groves BM, Amon KW. Accuracy of subxiphoid echocardiography for assessing left ventricular size and performance. Circulation 1980;61:367-373. [16.] Wallerson DC, Devereux RB. Reproducibility of echocardiographic left ventricular measurements. Hypertension 1987;9(suppl II):II-6-II-18. [17.] Lapido GOA, Dunn FG, Pringle TH, Bastian B, Lawrie TDV. Serial measurements of left ventricular dimensions by echocardiography. Assessment of week-to-week, inter- and intraobserver variability in normal subjects and patients with valvular heart disease. Br Heart J 1980;44:284-289. [18.] Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-310. [19.] Devereux RB. Detection of left ventricular hypertrophy by M-mode echocardiography. Anatomic validation, standardization, and comparison to other methods. Hypertension 1987;9(suppl II):II-19-II-26. [20.] Brenner JI, Waugh RA. Effect of phasic respiration on left ventricular dimension and performance in a normal population. An echographic study. Circulation 1978;57:122-127. From the Centre de Medecine Preventive Cardio-Vasculaire and the Departement de Biostatistiques et d'Informatique Medicale, Hopital Broussais, and the Service de Cardiologie, Hopital Saint Antoine, Paris, France. Manuscript received January 8, 1993; revised manuscript received and accepted March 18, 1993. Address for reprints: Eric Abergel, MD, Centre de Medecine Preventive Cardio-Vasculaire, Hopital Broussais, 96 rue Didot, 75674 Paris Cedex 14, France.