• Open Access

Left Atrial Ejection Fraction Assessed by Real-Time 3-Dimensional Echocardiography in Normal Dogs and Dogs with Myxomatous Mitral Valve Disease


  • Part of the study was presented as an abstract at the 21st ECVIM-CA Congress, September 2011, Seville, Spain

Corresponding author: A. Tidholm, Albano Animal Hospital, Danderyd, Sweden; e-mail: anna.tidholm@gmail.com



Real-time 3-dimensional (RT3D) echocardiography provides a novel technique for assessing left atrial ejection fraction (LAEF) in dogs.


Left atrial ejection fraction is associated with severity of myxomatous mitral valve disease (MMVD).


Privately owned dogs; 101 with MMVD and 52 healthy control dogs.


Prospective observational study using RT3D echocardiographic estimations of LA volumes at atrial end-diastole and atrial end-systole to calculate LAEF in comparison with conventional 2-dimensional echocardiographic variables.


Left atrial ejection fraction decreased with increasing LA to aortic ratio (LA/Ao), percentage increase in left ventricular (LV) internal dimension, corrected for body weight (BW), in diastole (LVIDd inc%) and systole (LVIDs inc%), and age for MMVD dogs, and with BW for control dogs. The final models in the multiple regression analyses included LVIDd inc% and age for MMVD dogs, and BW alone for control dogs. LAEF varied widely in both MMVD dogs and control dogs.

Conclusion and clinical importance

The wide variation of LAEF and the fact that LAEF does not appear to be an independent marker of disease severity suggest that the clinical importance of determining LAEF in dogs with MMVD might be limited.






aorta, aortic diameter in short-axis


congestive heart failure


coefficient of variation


ejection fraction


left atrium, left atrial diameter in short-axis


left atrial ejection fraction


left ventricle


left ventricular ejection fraction


end-diastolic left ventricular internal dimension

LVIDd inc%

percentage increase in end-diastolic left ventricular internal dimension


end-systolic left ventricular internal dimension

LVIDs inc%

percentage increase in end-systolic left ventricular internal dimension


myxomatous mitral valve disease


real-time 3-dimensional


left atrial volume at atrial end-diastole


left atrial volume at atrial end-systole

The left atrium (LA) modulates left ventricular (LV) filling through 3 major functions; it acts as a reservoir during ventricular systole (ie, atrial diastole), as a conduit during early ventricular diastole, and it has an active contractile function during late ventricular diastole (ie, atrial systole) delivering 15–30% of LV filling.[1] Accurate and reproducible measurements of LA volume and function are important for clinical decision-making in heart disease and have been shown to greatly influence outcome in a variety of human and canine heart diseases.[2-6]

Assessment of LA volume and contractile function can be made using two-dimensional (2D)[7, 8] and three-dimensional (3D)[9-11] echocardiography, computed tomography (CT),[12, 13] magnetic resonance imaging (MRI),[14] and velocity vector imaging (VVI).[15] 2D echocardiographic techniques generally underestimate LA volumes with no clear correlation with MRI or CT.[16] Real-time 3D (RT3D) echocardiography and CT have good agreement with the gold standard technique MRI in human studies, although RT3D systematically underestimates LA volumes.[9, 13, 17, 18] RT3D echocardiography was mildly superior to 2D using MRI and CT as gold standard for the estimation of LA volumes in healthy dogs.[19] Comparisons between RT3D and different 2D echocardiographic methods for estimation of LA size in dogs with myxomatous mitral valve disease (MMVD) did not show good correlation between 2D methods and the RT3D method without the use of allometric scaling equations.[20]

Although many studies in veterinary medicine concerning heart disease include assessment for LA size, few studies concern the function of the LA chamber itself. The objective of this study was to assess LAEF in dogs with and without MMVD and to assess potential associations between LAEF and a variety of clinical and echocardiographic variables.

Materials and Methods

Animals and Procedures

A total of 153 privately owned dogs, 101 dogs with MMVD, and 52 healthy control dogs, presented at Albano Animal Hospital, Stockholm, Sweden were included in the study. All dogs were examined using the same equipment and the same protocol. Control dogs were considered healthy based on normal findings on physical examination, ECG, echocardiography, and Doppler examinations. Diagnostic criteria for MMVD included thickened mitral valve leaflets and mitral regurgitation detected on color-coded Doppler echocardiogram.[21] Dogs with systemic disorders, congenital heart disease, or other acquired cardiovascular disorders including rhythm disturbances were excluded. All examinations were performed and evaluated by one veterinary specialist in cardiology (AT).

2- and 3-D Echocardiography

2D and RT3D echocardiographic examinations were performed with an ultrasound unit1 equipped with 3.0–8.5 MHz phased-array transducers (for 2D) and 7–2 MHz matrix transducer (for RT3D) with simultaneous ECG-recording in all dogs. Dogs were unsedated and gently restrained in left and right lateral recumbency for RT3D and 2D images, respectively, during the examination. Measurements of LA and Ao diameters were made on 2D parasternal short-axis view at the level of the Ao valve in early ventricular diastole at the 1st frame after Ao valve closure on a monitor freeze-frame image.[22] Measurements of LV diameters were made by 2D-guided M-mode according to the American Society of Echocardiography.[23] Ventricular end-diastole was defined as the 1st frame after mitral valve closure and ventricular end-systole was defined as the last frame before mitral valve opening. Percentage increase in LV internal diameter in diastole (LVIDd inc%) and in systole (LVIDs inc%) was calculated using the following formula: % increase = 100 × (observed dimension − expected normal dimension)/expected normal dimension. Expected normal dimensions according to body weight (BW) were calculated as follows: expected normal LVIDd = 1.53 × (BW)0.294; expected normal LVIDs = 0.95 × (BW)0.315.[24] Left ventricular ejection fraction (LVEF) was calculated using Teichholz formula,[25] as we have shown in an earlier study that there were no statistically significant differences among the Teichholz method, the Simpson's biplane method of disks, and the RT3D echocardiographic method for estimation of LVEF in dogs with and without MMVD.[26] Measurements were made on 3 consecutive cardiac cycles, and mean values were used in the statistical analyses.

RT3D images of LA were created using the 7–2 MHz matrix transducer to obtain a pyramidal volume in real time. Real-time volumes acquired from 4 to 7 cardiac cycles were obtained to produce a larger pyramidal volume, producing a full dataset. The lowest possible scan line density was used in all dogs to produce the largest possible pyramidal volume of acquisition.

RT3D Data Analysis

Off-line analyses of RT3D LA volumes were made by a software.2 Analysis of LA volume at atrial end-diastole (RT3DLAd) concurrent with ventricular end-systole was timed as the frame preceding mitral valve opening, and at atrial end-systole (RT3DLAs) concurrent with ventricular end-diastole timed as the 1st frame after mitral valve closure. The analysis involved manual definition of 4 reference points (2 points in each view) placed at the endocardial border of the LA side of the mitral annulus in both 4-chamber and 2-chamber views. A 5th reference point was placed at the midpoint of the dorsal LA border in either view. The endocardial border then was traced by an automated detection process to create a cast of the LA cavity (Fig 1). When the volume was computed, the endocardial border detection was verified for accuracy and manually edited when necessary. Three acquisitions were made for each variable, and mean values were used in the statistical analyses. RT3D-derived volumes of LA at atrial end-diastole and end-systole and the resulting LAEF were calculated according to the following formula: LAEF = 100 × ([RT3DLAd − RT3DLAs]/RT3DLAd). All measurements were made by 1 veterinary specialist in cardiology (AT).

Figure 1.

Real-time 3-dimensional image of left atrium (LA) volume at atrial end-diastole in a normal dog. After manual definition of 5 reference points, the endocardial border was automatically traced, and manually edited when necessary, to create a cast (in yellow) of the LA cavity.

Assessment of Variability

Within-day variability was assessed using 6 dogs, including 3 dogs without heart disease and 3 dogs with MMVD (Class BI and BII according to the ACVIM classification of cardiac disease).[27, 28] Each dog was examined 6 times on a given day. Each variable was measured on each of the 6 acquisitions for each dog, and the resulting mean values and standard deviations (SD) were used to determine the coefficient of variation (CV). All CV mean values were below 15%.

Statistical Analysis

A computer program3 was used for all statistical analyses. Data are presented as medians and interquartile ranges (IQR). The nonparametric Wilcoxon signed rank test was used for testing equality of medians between dogs with and without MMVD. Level of significance was set at < .05. Univariate and multiple regression analyses were used to evaluate associations between LAEF, and dog characteristics (age, sex, and BW), heart rate (HR) obtained from the echocardiogram, and echocardiographic measurements (LA/Ao ratio, LVIDd inc%, LVIDs inc%, and LVEF assessed by Teichholz formula). In the multiple regression modeling, analyses were performed in a backward stepwise manner, starting with all variables included in the model and then removing the variable with the highest P-value until all the remaining variables had a value of < .05. All variables were assessed only as main effects; no interaction terms were considered in the model. The adjusted R2 is defined as the percentage of the total sum of squares that can be explained by the regression, and it also considers the degrees of freedom for variables added. Level of significance was set at P < .05.


A total of 153 dogs of 58 breeds were included in the study: Cavalier King Charles Spaniel (16), mixed breed (13), New Foundland (8), Dachshund (10), Labrador Retriever (9), Miniature Schnauzer (6), Norfolk Terrier (6), Flat Coated Retriever (5), Staffordshire Bullterrier (5), and <5 dogs each of 49 other breeds. One hundred one dogs were diagnosed with MMVD and 52 dogs were healthy controls.

Seventy-three dogs were classified without congestive heart failure (CHF) (according to the ACVIM classification 68 dogs in Class B1 and 5 in B2), and 28 dogs were classified with CHF (class C). A total of 32 dogs underwent medical treatment where 28 dogs were treated with furosemide, 32 dogs were treated with pimobendan, 17 dogs were treated with benazepril, and 2 dogs were treated with spironolactone. Eighty-six dogs (56%) were males, and 67 dogs (44%) were females. Body weight ranged from 2.9 to 35 kg (11.2, IQR 8.1–17.4 kg) for MMVD dogs and from 2.8 to 55 kg (17, IQR 8.6–30.5 kg) in healthy control dogs. Age ranged from 3 to 18 years (9.6, IQR 7.8–11.5 years) for MMVD dogs and from 2 months to 14 years (3.8, IQR 1.5–6.8 years) for healthy control dogs. All dogs were in sinus rhythm, and HR ranged from 68 to 185 beats/min (125, IQR 100–140 beats/min) in MMVD dogs and from 67 to 190 beats/min (118, IQR 96–139 beats/min) in healthy control dogs (Table 1).

Table 1. Real-time 3-dimensional (RT3D) and 2-dimensional echocardiographic data in 153 dogs with (n = 101) and without (n = 52) myxomatous mitral valve disease (MMVD). Data are presented as median values and interquartile ranges (IQR)
VariableMMVD dogsControl dogsP-value
Sex (male/female)66/3520/32.0035
Age3.8 (1.5–6.8)9.6 (7.8–11.5)<.0001
Body weight11.2 (8.1–17.4)17 (8.6–30.5).009
Heart rate125 (100–140)118 (96–139).43
LA/Ao1.3 (1.1–1.5)1.1 (1–1.2)<.001
RT3DLAd (mL/kg)1.1 (0.8–1.8)0.6 (0.5–0.8)<.0001
RT3DLAs (mL/kg)0.6 (0.4–1.1)0.4 (0.3–0.5)<.0001
RT3DLAEF%41.3 (31.9–52.7)37.5 (29–48.8).08
LVIDd inc%15 (4–32)2 (−10–11)<.0001
LVIDs inc%3 (−10–20)0.1 (−10–13).19
LVEF%70.8 (62.1–79.3)63.3 (54.8–68.3).65
Mitral E m/s0.83 (0.67–1.1)0.68 (0.6–0.83)0.0001
Mitral A m/s0.69 (0.59–0.83)0.53 (0.47–0.66)<0.0001
Mitral E/A1.2 (1.1–1.5)1.3 (1.1–1.5)0.65

2-D and 3D Echocardiography

For all dogs, RT3DLAd by BW ranged from 0.3 to 5.8 mL/kg (0.86 mL/kg, IQR 0.61–1.3) and RT3DLAs by BW ranged from 0.16 to 3.9 mL/kg (0.45 mL/kg, IQR 0.34–0.74). LAEF ranged from 8 to 74% in all dogs, with no overall significant difference (= .08) between dogs with and without MMVD (38%, IQR 29–49 and 41%, IQR 32–53, respectively). MMVD dogs had significantly greater values for LA/Ao, RT3DLAd, and RT3DLAs by BW, LVIDd inc%, and mitral E and A wave velocity compared with normal dogs, whereas there was no significant difference between groups concerning LAEF, LVIDs inc%, LVEF, and mitral E/A (Table 1). Seven (5%) dogs had an E/A > 2 and 22 (14%) dogs had an E/A < 1. LVEF < 45% was found in 5 (3%) dogs in this study. Thirty-three (22%) of the dogs had LA/Ao > 1.4.[2, 22, 29]

Univariate Regression Analysis

LAEF was negatively associated with increasing LA/Ao (R2 = 0.13, = .0003), LVIDd inc% (R2 = 0.15, < .0001), LVIDs inc% (R2 = 0.08, P = .004) and age (R2 = 0.07, = .006) for MMVD dogs, and with BW (R2 = 0.27, < .0001) for control dogs. There was no association between LAEF and sex (= .4), HR (= .12), mitral E/A (= .06) or LVEF (= .36) assessed by the Teichholz method.

Multiple Regression Analysis

Of the 8 included variables, LVIDd inc% and age remained significant in the multiple regression analysis for MMVD dogs, and only BW for control dogs. The adjusted R2-values were 0.2 for age (= .0022) and LVID inc% (P = .0002) in MMVD dogs, and 0.27 for BW (< .0001) in control dogs.


In the present study, LAEF was negatively associated with increasing LV volumes in systole and diastole in MMVD dogs, suggesting that LV volume overload reduces LA contractile function. A reduction in LA contractile performance has been reported in human patients with heart failure caused by increased LV filling pressures.[30] LAEF was negatively associated with LV systolic and diastolic dysfunction in human patients, and a progressive decline in LAEF was demonstrated in concert with deteriorating diastolic dysfunction.[11] The increased LA pressure, ie, LA afterload, resulting from an increased LV chamber stiffness and wall stress as well as increased LA filling might lead to LA dilatation and reduced systolic performance.[31] In the present study, there was no association between LAEF and mitral E/A or LVEF (assessed by Teichholz method). However, overt diastolic LV dysfunction was not a feature in the dogs in this study, as only 7 (5%) dogs had an E/A > 2 and 22 (14%) dogs had an E/A < 1. Systolic dysfunction, defined as LVEF < 45%,[26] was found in only 5 (3%) of dogs in this study.

LA contractile function has been assessed using 2D methods, such as LA fractional area change, LA shortening fraction, or LAEF, using 1- or 2-dimensional measurements.[5, 6, 8, 31, 32] However, 1- and 2-dimensional methods, although easily acquired, appear to be inaccurate to quantitate LA size as previously shown by our group and others.[16, 20, 33] In this study, LAEF decreased with increasing LA and LV volume overload in the absence of overt LV diastolic dysfunction. Several human studies have shown a decline in LAEF with worsening LV dysfunction.[11, 34] The decrease in LAEF associated with LA and LV volume overload may contribute to a decrease in stroke volume and cardiac output.

Left atrial EF was negatively associated with increasing size of LA for MMVD dogs, indicating that stretching of the myocytes decreases their ability to contract. The Frank–Starling mechanism predicts an initial increase in contractile function as LA enlarges. However, as LA dilatation, defined as LA/Ao > 1.4,[2, 22, 29] was present in only 22% of the dogs, the study population is not suitable for evaluating the Frank–Starling mechanism in MMVD dogs with LA dilatation. LA dilatation beyond the peak length-action curve appears to be associated with LA contractile dysfunction, as LAEF is negatively correlated with increased LA size in human patients with moderate to severe LV systolic dysfunction.[35] There is a positive correlation between LAEF and decreasing LA volume after mitral valve repair in human patients,[36, 37] whereas LAEF is not significantly correlated with LA size in patients with myocardial infarctions.[16] Increased LAEF preceded LA dilatation in a study of human patients with early stage of hypertension.[38]

LAEF was negatively correlated with age in MMVD dogs both in the univariate and in the multiple regression analysis in this study. A decrease in LAEF with increasing age along with an increase in LA end-diastolic volume has also been reported in healthy humans.[10, 39] These findings might be considered consistent with a physiologic decrease in LV compliance and increased LA afterload that occur with increasing age.[40]

LAEF was negatively correlated with BW in control dogs in this study. As BW differed between breeds, this observation might be caused by differences in breed rather than in BW in itself. Heart rate, which has been considered to be lower in large breeds compared with small breeds,[41, 42] although recent studies appear to contradict this hypothesis,[43, 44] did not correlate with LAEF in this study. Whether or not LAEF differs among breeds remains to be investigated.

In this study, LAEF varied widely in both MMVD dogs and in normal control dogs with medians of approximately 40%. A substantial variation in LAEF has been reported in human patients with various heart diseases,[9, 12] and in healthy subjects.[10, 11, 45] The contractile performance of LA might thus not be used as a hallmark of cardiac disease in our study. The wide variation of LAEF and the fact that LAEF does not appear to be an independent marker of disease severity in MMVD suggest that the clinical usefulness of LAEF in dogs with MMVD might be limited.

Limitations of the Study

The MMVD dogs were unevenly distributed among the different classes of heart failure with a majority (73%) of dogs classified as B1, precluding group-wise comparison, and evaluation of effect of CHF and severe LA dilatation on LAEF. The fact that only few (19%) of the dogs showed evidence of LV diastolic dysfunction precluded evaluation of effect of diastolic function on LAEF. A possible influence of cardiac medication on LAEF could not be evaluated in the present study because only 32 of 153 (21%) dogs received cardiac medications at the time of enrollment and because of variations in the combinations and dosages of cardiac medications administered to these dogs.

In conclusion, LAEF decreased with increasing LA and LV volume overload. LA contractile performance decreased with increasing age in MMVD dogs, and with increasing BW in control dogs. LAEF varied widely in both MMVD dogs and in control dogs, and as LAEF does not appear to be an independent marker of disease severity, its usefulness in clinical practice may be questioned.


Conflict of Interest Declaration: Authors disclose no conflict of interest.


  1. 1

    IE33, Philips Ultrasound, Bothell, WA

  2. 2

    QLAB advanced quantification system version 5.0, Philips Ultrasound

  3. 3

    JMP, v.5.1, SAS Institute Inc, Cary, NC