Echocardiographic Quantitation of Left Ventricular Size and Function in Indian Mongrel Dogs

Two-dimensional echocardiography provides real-time imaging of cardiac structures, cross-sectional information and numerous imaging planes for complete cardiac evaluation (O'Grady et al., 1986), therefore it is the preferred noninvasive modality used for quantitation of cardiac chamber size and function. Two-dimensional echocardiographic images have close resemblance to the gross anatomic appearance of cardiac structures and, therefore provide more useful information about change in shape, size and function of heart in various disease conditions (Visser et al., 2019).
       
Normal two-dimensional echocardiographic values of left ventricular chamber, interventricular septum and left ventricular posterior wall dimensions and systolic functional indices are required for comparison and evaluation of dogs suspected of having heart diseases (O'Grady et al., 1986). Unlike human  medicine, recommended standards for quantitation of cardiac chamber size and function by  two-dimensioanl echocardiography in dogs do not currently exist (Lang et al., 2015). Available  veterinary literature  proposing reference intervals for linear measurements of left ventricular size and function in healthy dogs has been derived using either small number of dogs or single breed (Hansson et al., 2002; Rishniw and Erb, 2000) except few studies in which large sample size and different dog breeds  were used (O'Grady et al., 1986; Visser et al., 2019).
       
To the author¢s knowledge, no work on cardiac chamber quantitation using two-dimensional echocardiography is reported in Indian mongrel dogs till date. The present study envisaged quantitative assessment of left ventricular size and function in healthy mongrel dogs using two-dimensional echocardiographic measurements obtained from different tomographic planes with the objectives  to standardize and generate reference data of linear and cross-sectional measurements of left ventricle size and function and to study the possible correlationship of data obtained with body weight.

The present study was conducted in the session 2021-22 and 2022-23 at the Department of Veterinary Surgery and Radiology, Dr. GC Negi College of Veterinary and Animal Sciences, Chaudhary Sarwan Kumar Himachal Pradesh Krishi Vishvavidyalaya, Palampur as a part of first authors MVSc thesis. Fifty client owned healthy mongrel dogs (25 male; 25 female) presented for routine health check-up, vaccination and elective surgeries were the subjects of study. Selection criteria for clinically healthy dogs involved thorough assessments, including clinical evaluation, hematological analyses, radiological examination, electrocardiographic and routine echocardiographic evaluation. Dogs deemed healthy after  preliminary examination underwent two-dimensional echocardiographic examination. A verbal consent was obtained from each dog owner prior to imaging studies.
       
The dogs were restrained manually on a specially designed table with a cut-out hole. The thorax was prepared on both sides sides from 3^rd-7^th rib and 1-5 cm lateral to sternum for right and left parasternal echocardiographic examination. Echocardiography was performed with an ultrasound machine, ACUSON X300 Premium Edition Diagnostic Ultrasound System, Siemens Medical Solutions USA, Inc. equipped with a 4.0-8.0 MHz multi-frequency phased array probe. Simultaneous electrocardiogram was recorded and superimposed on image display. Gain settings and gray scales were adjusted for better imaging of endocardial and epicardial surfaces. The vedio-tape recorded images were later replayed and measured using software available in ultrasound machine. Two-dimensional linear and area measurements of left ventricle was performed in right parasternal and left parasternal imaging planes (Thomas et al., 1993) using the methods described by O'Grady et al., (1986). Same standardized imaging protocol was followed for each examination. In right parasternal long axis four-chamber imaging plane, care was taken to avoid left ventricle foreshortening. Left ventricle chamber dimension and wall thickness were measured in end-diastolic and end-systolic frames using blood-tissue interface (ie, inner edge-to-inner edge measurement technique). Following measurements of left ventricle were obatined in different tomographic planes.
1.     Left ventricular internal length in diastole (L1d) and systole (L1s) was obtained from right parasternal long axis (RPLAx) view optimized for  the left ventricular apex and mitral valve. L1 was measured from the midpoint of the left ventricular side of the mitral annulus extending to the apical endocardium (Fig 1a). Left ventricle internal length in diastole (L2d) and systole (L2s) was obtained from left parasternal apical long axis (LPALAx) optimized for left ventricle and mitral valve. L2 was measured from the midpoint of a reference line connecting the left ventricular side of the mitral annulus and extending to the apical endocardium (Fig 1b).
 

 
2.     Left ventricular internal dimension, interventricular septum and free wall thickness in diastole and systole was measured in RPLAx view optimized for interventricular septum, left ventricular outflow tract, mitral valve and left ventricular free wall (excluding papillary muscle images) (O’Grady et al., 1986) (Fig 2a).
 

 
3.     Left ventricular internal dimension, interventricular septum and free wall thickness in diastole and systole was measured in RPSAx view optimized for chordate tendinae level of left ventricle (Fig 2b) (O’Grady et al., 1986).
4.     Left ventricular internal and external area was measured from RPSAx view at the left ventricular high papillary muscle level. Internal area was measured by tracing the trailing edge echo of the left ventricular septal endocardium and leading-edge echo of the left ventricular free wall endocardium in end-diastole.External area was measured by tracing the trailing edge echo of the right ventricular septal endocardium and the leading-edge echo of the pericardium in end-diastole (Fig 3).
 

 
5.     Mitral valvular orifice area was measured from RPSAx view of mitral valve in the frame showing maximal opening of anterior and posterior mitral leaflets (Fig 4).
 

 
Following systolic function indices were derived:
1.     Fractional shortening, FS (%), calculated from left ventricular internal dimensions in RPLAx and RPSAx using Teichloz formula (Teichloz et al., 1976).
   
where:
LVIDd: Left ventricular internal dimension at end-diastole. LVIDs: Left ventricular internal dimension at end-systole.
2.     Ejection fraction, EF (%), derived from left ventricular diastolic and systolic volumes obtained using monoplane Simpson¢s method of discs (SMOD) in RPLAx.
 
Statistical analysis
 
Data was analyzed by IBM SPSS Statistics 29.0 software. Bivariate Pearson¢s correlation test was used to establish body weight correlation with echocardiographic parameters. Linear regression analysis was performed to assess relationship between echocardiographic parameters and body weight and, between same measurements obtained from different tomographic planes. Linear regression equation Y=bX+a was utilized to model and quantify relationship between dependent echocardiographic parameter (Y) and independent variable body weight (X). One-way analysis of variance (ANOVA) was used to compare mean values among males and females.

Two-dimensional linear and cross-sectional measurements of left ventricle,interventricular septum, left ventricular free wall and mitral valve was standardized in 50 healthy mongrel dogs ageing 6 months-5 year (mean age= 1.59± 0.13 yr) and weighing 7-30 kg (mean body weight=16.66±0.73 kg).Two-dimensional echocardiographic images obtained in study resembled images documented by Thomas (1984) and O’Grady et al., (1986).Right parasternal location provided superior images compared to left parasternal, similar to findings of O’Grady et al., (1986). When measuring left ventricle dimensions, it was found that systolic and diastolic dimensions in short axis views were consistently greater compared to long-axis views (Table 1), aligning with the findings of O’Grady et al., (1986).
 

 
Left ventricular internal lengths
 
Left ventricular internal lengths in diastole and systole in RPLAx (L1d and L1s) and LPALAx (L2d and L2s) (Table 1) were lower than values (58.16±10.82 and 44.86±8.38 mm) and (51.74±10.42; 39.68±8.89 mm), respectively reported by O’Grady et al., (1986) in normal dogs.
 
Left ventricular internal dimensions (RPLAx)
 
Diastolic interventricular septum thickness (IVSd) (Table 1) was greater than value given by O’Grady et al., (1986) (6.91± 1.65 mm) but lower than values reported by Oliveira et al., (2014) (10.1±1.1 mm). Systolic interventricular septum thickness (IVSs) was near to value reported by O’Grady et al., (1986) (10.08±2.43 mm) but lower than value reported by Oliveira et al., (2014) (13.7±1.9 mm). Diastolic left ventricular internal dimension (LVIDd) was near to value reported by O’Grady et al., (1986) (33.92± 5.54 mm) but lower than value reported by Oliveira et al., (2014) (38.80±4.00 mm). Systolic left ventricular internal dimension (LVIDs) was  lower than values reported by O’Grady et al., (1986) (24.64±4.79 mm) and Oliveira et al., (2014) (27.7± 3.10 mm) in normal dogs. Diastolic left ventricular free wall thickness (LVFWd) was slightly lower than value reported by Oliveira et al., (2014) (9.90±1.30 mm) but greater than value reported by O’Grady et al., (1986) (7.88± 2.16 mm). Systolic left ventricular free wall thickness (LVFWs) was slightly greater than value reported by O’Grady et al., (1986) (10.85±2.25 mm) but lower than value given by Oliveira et al., (2014) (14.80±2.00 mm).
 
Left ventricular internal dimensions (RPSAx)
 
Diastolic interventricular septum thickness (IVSd) was near to value reported by Oliveira et al., (2014) (9.5± 0.80 mm) but greater than value reported by O’Grady et al., (1986) (8.26±1.97 mm).  Systolic interventricular septum thickness (IVSs) was near to value reported by O’Grady et al., (1986) (10.19±2.15 mm) but lower than that given by Oliveira et al., (2014) (12.8± 0.06 mm). Diastolic and systolic left ventricular internal dimensions (LVIDd and LVIDs) in RPSAx were near to values reported by O’Grady et al., (1986) (34.55±1.97 mm and 25.89±4.77 mm) but lower than value given by Oliveira et al., (2014) (40.0±4.3 mm and 28.9±3.40 mm).Diastolic left ventricular free wall thickness (LVFWd) was near to value reported by Oliveira et al., (2014) (9.7±0.11 mm) but slightly greater than value reported by O’Grady et al., (1986) (8.01±2.32 mm). Systolic left ventricular free wall thickness (LVFWs) was lower than value reported by Oliveira et al., (2014) (14.3±0.16 mm) but near to value reported by O’Grady et al., (1986) (10.33±2.77 mm).
 
Left ventricular internal and external area and mitral valve orifice area (RPSAx)
 
 Internal and external area of left ventricle and mitral valve orifice area (Table 1) were slightly lower than values reported by O’Grady et al., (1986) (9.17±2.17 cm², 22.85±7.6 cm2 and 3.69±69 cm², respectively).
 
Left ventricular systolic function indices (Fractional shortening and ejection fraction)
 
Fractional shortening (FS) in RPLAx was slightly greater than value reported by O’Grady et al., (1986) (28 %; 21-34%) and Oliveira et al., (2014) (28.4 %; 26.89-29.94%) while in RPSAx, FS was near to values  reported by O’Grady et al., (1986) (26%; 14-39 %) and (Oliveira et al., 2014) (28.1 %; 26.52-29.68%).Visser et al., (2019) reported a cut-off value of FS of 22 % suggesting that  many apparently large breed healthy dogs can have a FS close to 20% for years without a known cardiac event. Ejection fraction in RPLAx (53 %; 45-72%) was within the reference range proposed by O’Grady et al., (1986) (54 %; 42-64%) and Visseret_al(2019) (46-82.5%).
 
Correlation of two-dimensional echocardiographic measurements with body weight
 
Two-dimensional echocardiographic measurements showed significant (P<0.01) positive linear correlationship with body weight (Table 2), concurring with previous findings in normal dogs (O'Grady et al., 1986; Muzzi et al., 2006; Visser et al., 2019) and with those of left atrium and left ventricle size in cats (Haggstorm et al., 2016; Karsten et al., 2017). Significant (P<0.001) positive correlation between body weight and diastolic and systolic left ventricular length (L1d, LIs) in RPLAx and LPALAx (L2d, L2s) was in consonance with the findings of O’Grady et al., (1986). Significant (P<0.001) positive correlation between body weight and diastolic and systolic interventricular septum thickness (IVSd, IVSs), left ventricular internal dimension (LVIDd, LVIDs) and left ventricular free wall thickness (LVFWd, LVFWs) in right parasternal long and short axis views was in agreement with findings of other investigators (O'Grady et al., 1986; Oliveira et al., 2014 and Visser et al., 2019). Significant (P<0.001) positive correlation between body weight and left ventricular internal and external area and mitral valve orifice area was in line with findings of O’Grady et al., (1986).
 

       
Significant weak negative correlation between body weight and FS (RPLAx, RPSAx) observed during this study (Table 2) substantiate the findings of Kayar et al., (2006), Oliveira et al., (2014) and Visser et al., (2019) but negate the findings of other investigators who reported no correlation between FS and body size, suggesting that FS is independent of body weight (Cornell et al., 2004; Bodh et al., 2019). Visser et al., (2019 observed significant weak negative correlations of EF and fractional area change with body weight suggesting that larger dogs might exhibit relatively decreased systolic function compared to smaller dogs owing to their higher resting vagal tone, calmer demeanor and increased athleticism. Variation in the observation of FS and EF might result due to dependence of these parameters on multiple factors such as preload, afterload and contractility, which influence them (O’Leary et al., 2003). In general, linear measurements are linearly related to body length or body weight^1/3 and area and volume measurements are linearly related to body weight^2/3. Few studies reported a non-linear relationship between body weight and linear measurements of cardiac chamber size in dogs (Cornell et al., 2004; Hall et al., 2008) and cats (Scansen and Morgan, 2015). Contrary to above findings, our study supports the notion that linear cardiac measurements in dogs are indeed linearly related to body weight.
 
Effect of gender on two-dimensional echocardiographic measurements
 
Diastolic left ventricular internal length (Ld; RPLAx), systolic interventricular septum thickness (IVSs, RPLAx), systolic left ventricular internal dimension (LVIDs, RPSAx) and diastolic left ventricular free wall thickness (LVFWd, RPSAx) and internal and external area of left ventricle (RPSAx) showed significant (P<0.05) gender-based differences, with male dogs having greater values than females corroborating earlier reports (Muzzi et al., 2006; Bavegems et al., 2007). Mitral valve orifice area was significantly (P<0.05) greater in female dogs compared to males, substantiating the findings of Bodh et al., (2019) who reported significantly (P<0.05) higher mitral valve excursion amplitude in Indian Spitz dogs. Significantly (P<0.05) greater LVIDs (RPSAx) in male dogs was in line with finding of Muzzi et al., (2006) but contrary to findings of Bavegems et al., (2007) who reported significantly larger values of left ventricle internal diameters in female Whippets, attributable to their greater mean heart weight to body weight ratio compared to males. Significantly (P<0.05) greater LVFWs (RPSAx) in male dogs was in line with findings of Muzzi et al., (2006) who attributed increased thickness of left ventricular free wall due to work hypertrophy and higher body of males dogs. Other parameters viz. L2d and L2s (LPALAx), IVSd and IVSs (RPSAx), LVIDd and LVIDs (RPLAx), LVFWd and LVFWs (RPLAx), were unaffected by gender. Non-significant effect of gender on FS and EF supported previous findings in dogs (Muzzi et al., 2006; Visser et al., 2019) and cats (Haggstorm et al., 2016; Karsten et al., 2017).
 
Correlation between structures imaged and measured from different tomographic planes
 
All two-dimensional linear and cross-sectional measurements of left ventricle measured from the right parasternal long axis, right parasternal short axis, left parasternal long axis and left parasternal apical long axis imaging planes displayed significant positive correlations (P<0.01) with each other (Table 3) consistent with findings reported by O’Grady et al., (1986) in normal dogs.
 

All linear and area measurements of left ventricle correlated positively and significantly (P<0.001) with body weight while fractional shortening and ejection fraction showed significant (P<0.001) negative correlation with body weight. Majority of parameters exhibited non-significant effect of gender except diastolic left ventricular internal length (L1d, RPLAx),systolic interventricular septum thickness (IVSs, RPLAx), systolic left ventricular internal dimension (LVIDs, RPSAx), diastolic left ventricular free wall thickness (LVFWd, RPSAx) that were significantly (P<0.01) higher in male dogs. All linear and area measurements obtained from different tomographic planes correlated positively and significantly (P<0.001) with each other.Reference values of left ventricular size and function generated in this study will aid in cardiac disease diagnosis in Indian mongrel dogs.

The authors would like to thank Department of Veterinary Medicine and Veterinary Clinical Complex for the referrals.

The authors declare that they have no conflict of interest.