Summary
Background
Four two-dimensional echocardiographic methods (cube, ellipsoid, Simpson’s and area-length) can be used to assess left atrial volume (LAV).
Aims
To compare absolute LAV measurements and evaluate agreement regarding the semiquantitative assessment of degree of left atrial (LA) enlargement, between methods.
Methods
We prospectively measured LAV in 51 healthy volunteers using the four methods, and defined thresholds for moderate (mean + 2 standard deviations [SDs]) and severe (mean + 4 SDs) LA enlargement for each method. In 372 patients referred for echocardiography, we compared absolute LAV measurements and agreement between methods.
Results
LAV was significantly different between methods in the healthy volunteer group (11 ± 4, 17 ± 3, 26 ± 6 and 28 ± 7 mL/m 2 , respectively; P < 0.0001), resulting in different thresholds for moderate and severe LA enlargement. LAV was also significantly different in the 372 patients (30 ± 20, 47 ± 27, 61 ± 34 and 65 ± 36 mL/m 2 , respectively; P < 0.0001). Agreement regarding degree of LA enlargement (none, moderate, severe), using the area-length method as reference, was modest with the cube method (kappa = 0.41), correct with the ellipsoid method (kappa = 0.60) and excellent with Simpson’s method (kappa = 0.83).
Conclusion
The choice of the method had a major effect on assessment of degree of LA enlargement. Our results suggest that the cube and ellipsoid methods, which significantly underestimated LAV and provided modest agreement, should be disregarded. In contrast, Simpson’s method and the area-length method were slightly different, but showed close agreement, and should be preferred, using dedicated thresholds (50 and 56 mL/m 2 respectively).
Résumé
Contexte
Il existe quatre méthodes de mesure de l’oreillette gauche (OG) en échocardiographie transthoracique (cube, ellipse, Simpson’s, aire/longueur).
Objectif
Évaluer l’impact du choix de la méthode de mesure du volume de l’OG en échocardiographie sur l’estimation de son degré de dilatation.
Méthodes
Nous avons calculé prospectivement le volume de l’OG avec chaque méthode chez 51 sujets sains, ainsi que les valeurs seuil permettant de définir une OG dilatée (moyenne + 2DS), et très dilatée (moyenne + 4DS). Nous avons comparé dans une cohorte de 372 patients les valeurs absolues de volume de l’OG et la concordance des quatre méthodes de mesure.
Résultats
Le volume OG était différent selon la méthode utilisée (11 ± 4, 17 ± 3, 26 ± 6 et 28 ± 7 mL/m 2 , respectivement ; p < 0,0001), avec en conséquence des valeurs seuil différentes pour définir une dilatation modérée ou sévère. Le volume OG était également différent pour les patients (30 ± 20, 47 ± 27, 61 ± 34 et 65 ± 36 mL/m 2 , respectivement ; p < 0,0001). La concordance pour définir le degré de dilatation OG (nulle, modérée, sévère) entre la méthode aire/longueur et la méthode du cube était modeste (kappa = 0,41), correcte avec la méthode de l’ellipse (kappa = 0,60) et excellente avec la méthode du Simpson (kappa = 0,83).
Conclusion
Le choix de la méthode influence fortement l’appréciation du degré de dilatation de l’OG. Les méthodes aire/longueur et Simpson, contrairement à celle du cube et de l’ellipse, donnaient des volumes peu différents et une excellente concordance. Elles devraient être privilégiées dans l’évaluation de la dilatation OG en utilisant des seuils dédiés (50 et 56 mL/m 2 respectivement).
Background
Left atrial (LA) size is an important prognostic marker in various cardiovascular diseases, such as valvular heart disease, hypertension, dilated cardiomyopathy, hypertrophic cardiomyopathy and stroke . Transthoracic echocardiography is the most widely used method available to assess LA size, and the superiority of LA volume (LAV) over LA diameter is now well established . Thus, LAV should be measured consistently during each echocardiogram. However, LAV can be calculated using different methods. The most accurate method remains a matter of debate, and the effect of the method on the assessment of the degree of LA enlargement has never been evaluated.
Thus, in this prospective study, we aimed: to compare LAV measurements between four methods (the cube method, the ellipsoid method, the biplane Simpson’s method and the biplane area-length [A/L] method) to define specific thresholds for moderate and severe LA enlargement for each method, in a healthy volunteer group; and to evaluate the agreement between methods regarding the semiquantitative evaluation of the degree of LA enlargement, in a subset of patients referred for clinically indicated transthoracic echocardiography.
Methods
Population
We prospectively enrolled two different groups of participants: healthy volunteers with no history of cardiovascular disease (nurses, physicians and medical students enrolled in an ongoing prospective study [GENERAC, clinicalTrial.gov number NCT00647088 ]); and consecutive patients who underwent transthoracic echocardiography between January and December 2010 conducted by the last author (D.M.-Z.) using an IE33 ultrasound system (Philips, Amsterdam, Netherlands). Exclusion criteria were a poor echocardiographic window or incomplete data for the measurement of LAV by the four methods. All participants gave informed consent.
Clinical data
For each volunteer or patient, weight, high, body surface area (BSA), medical history and indication for echocardiography were collected. BSA was calculated as 2 × √([weight (kg) × height (m)]/3600) and body mass index as (weight [kg]/height 2 [cm]).
Echocardiography
All participants underwent comprehensive echocardiography prospectively, conducted by the same experienced physician (the last author, D.M.-Z.). All measurements were performed in end-systole just before mitral valve opening. LA anteroposterior diameter (D) was measured in parasternal long-axis view. Other measurements for LAV calculations were performed in the apical three- and four-chamber views using the zoom. The mediolateral diameter (D1), the area (A1) and the vertical length between the annulus and the posterior LA wall (L1) were measured in the four-chamber view. The area (A2) and the vertical length between the annulus and the posterior LA wall (L2) were measured in the apical three-chamber view ( Fig. 1 ).
LAV was calculated using the four methods. The formula for the cube method is: LAV = [π/6] × D 3 . The formula for the ellipsoid method is LAV = π (D × D1 × L1)/6. In the biplane Simpson’s method of disks, the left atrium (LA) is divided into a pile of disks perpendicular to the longitudinal length direction, in both the four-chamber and three-chamber views. The radius of each disk is measured from the longitudinal axis to the LA contour in the two perpendicular planes. The volume of each disk is calculated automatically, and LAV is calculated by the summation of the disk’s volume: LAV = π/4Σ(i = 1 to 20) ai × bi × L/N, where ai and bi are 20 discs obtained in the two orthogonal incidences (the four- and three-chamber views). The formula for the biplane A/L method is LAV = (8/3) π (A1 × A2)/([L1 + L2]/2)
Importantly the same tracing was used to perform all measurements (mediolateral diameter; vertical lengths L1 and L2; and LA areas A1 and A2) and to calculate LAV using the method of disks, to ensure that differences between methods were only due to the method and not to the tracing. Results were indexed to the BSA (LAV index [LAVI]).
Statistical analysis
Continuous variables are expressed as mean ± standard deviation (SD), median [95% confidence interval] or number (percentage) of patients. Comparisons of LAVI measured by the different methods were made using one-way analysis of variance or a paired t test, as appropriate. For each method, moderate LA enlargement was defined as mean LAVI measured in the healthy population + 2 SDs; severe LA enlargement was defined as mean LAVI + 4 SDs. The agreement between methods regarding the semiquantitative assessment of the degree of LA enlargement in the patient group was evaluated using the kappa value. The A/L method – previously reported as providing the most accurate assessment of LAV compared with computed tomography – was used as reference. When the degree of LA enlargement was similarly graded between two methods, they were considered concordant. In contrast, if the degree of LA enlargement was different, methods were considered discordant (with over- or underestimation compared with the A/L method). A P value < 0.05 was considered statistically significant. Statistics were performed using JMP software (SAS, Cary, NC, USA).
Results
Population
Fifty-one healthy volunteers and 372 patients were prospectively enrolled. Clinical and echocardiographic characteristics are presented in Table 1 . Mean age in the healthy volunteer group was 33 ± 12 years and 22 (43%) were men. Left ventricular ejection fraction and systolic pulmonary artery pressure were normal by design. Mean age in the patient group was 63 ± 17 years, 204 (55%) were men, left ventricular ejection fraction was normal in 336 (90%) patients and 54 (15%) had an elevated systolic pulmonary artery pressure (>40 mmHg). Two-hundred and eighty-three (76%) patients were referred for evaluation of valvular heart disease and the remaining patients ( n = 89; 24%) were referred for other reasons (ischaemic, dilated or hypertrophic cardiomyopathy and atrial fibrillation).
| Healthy volunteers ( n = 51) | Patients ( n = 372) | |
|---|---|---|
| Age (years) | 33 ± 12 (28 [21–54]) | 63 ± 17 (66 [25–89]) |
| Men | 22 (43) | 204 (55) |
| Body mass index (kg/m 2 ) | 23 ± 3 | 26 ± 5 |
| Atrial fibrillation | 0 (0) | 68 (18) |
| Normal ejection fraction | 51 (100) | 336 (90) |
| Systolic pulmonary artery pressure (mmHg) | 21 ± 4 (20 [14–32]) | 41 ± 14 (38 [22–74]) |
| Left atrial diameter (mm) | 34 ± 4 (36 [26–42]) | 45 ± 9 (44 [28–66]) |
| Indexed left atrial diameter (mm/m 2 ) | 19 ± 2 (19 [15–23]) | 25 ± 6 (24 [16–38]) |
| Indexed left atrial volume (mL/m 2 ) | ||
| Cube method | 11 ± 4 (11 [4–21]) | 30 ± 20 (25 [7–83]) |
| Ellipsoid method | 17 ± 3 (17 [9–25]) | 47 ± 27 (41 [16–122]) |
| Biplane Simpson’s method of disks | 26 ± 6 (25 [13–44]) | 61 ± 34 (53 [22–147]) |
| Biplane area/length method | 28 ± 7 (26 [14–47]) | 65 ± 36 (56 [23–158]) |
Left atrial volume in healthy volunteers
In the 51 healthy volunteers, mean LAVI was significantly different using the four different methods: 11 ± 4 mL/m 2 using the cube method, 17 ± 3 mL/m 2 using the ellipsoid method, 26 ± 6 mL/m 2 using the biplane Simpson’s method and 28 ± 7 mL/m 2 using the biplane A/L method ( P < 0.0001) ( Table 1 ). Mean LAVI measured using the biplane Simpson’s method and the biplane A/L method was also significantly different ( P < 0.0001). Threshold values for defining moderate and severe LAVI enlargement for each method are presented in Table 2 . For example, using the biplane A/L method, LAVI was moderately enlarged above 42 mL/m 2 and severely enlarged above 56 mL/m 2 .
