Echocardiographic assessment of right ventricular systolic function in conscious healthy dogs: Repeatability and reference intervals

Journal of Veterinary Cardiology (2015) 17, 83e96

www.elsevier.com/locate/jvc

Echocardiographic assessment of right ventricular systolic function in conscious healthy dogs: Repeatability and reference intervals Lance C. Visser, DVM, MS , Brian A. Scansen, DVM, MS*, Karsten E. Schober, DVM, PhD , John D. Bonagura, DVM, MS Department of Veterinary Clinical Sciences, College of Veterinary Medicine, The Ohio State University, 601 Vernon L. Tharp Street, Columbus, OH 43210, USA Received 31 July 2014; received in revised form 29 September 2014; accepted 21 October 2014

KEYWORDS Canine; Cardiac; Reference range; Reproducibility; Right ventricle

Abstract Objectives: To determine the feasibility, repeatability, intra- and interobserver variability, and reference intervals for 5 echocardiographic indices of right ventricular (RV) systolic function: tricuspid annular plane systolic excursion (TAPSE), fractional area change (FAC), pulsed wave tissue Doppler imagingderived systolic myocardial velocity of the lateral tricuspid annulus (S’), and speckle-tracking echocardiography-derived global longitudinal RV free wall strain and strain rate. To explore statistical relationships between RV systolic function and age, gender, heart rate, and bodyweight. Animals: 80 healthy adult dogs. Methods: Dogs underwent 2 echocardiographic examinations. Repeatability and intra-observer and inter-observer measurement variability were quantified by average coefficient of variation (CV). Relationships between RV function and age, heart rate and bodyweight were estimated by regression analysis. Results: All indices were acquired with clinically acceptable repeatability and intra- and inter-observer variability (CVs < 10%). No differences were identified between male and female dogs. Allometric scaling by bodyweight demonstrated significant, clinically relevant correlations between RV function and bodyweight (all p
0.001) as follows: TAPSE e strong positive correlation (r2 ¼ 0.75); S’ e moderate positive correlation (r2 ¼ 0.31); strain rate e moderate negative correlation

Presented in abstract form as an oral presentation at the 2014 ACVIM Forum, Nashville, TN. * Corresponding author. E-mail address: [email protected] (B.A. Scansen). http://dx.doi.org/10.1016/j.jvc.2014.10.003 1760-2734/ª 2014 Elsevier B.V. All rights reserved.

84

L.C. Visser et al. (r2 ¼ 0.44); FAC and strain e weak negative correlations (r2 ¼ 0.22 and 0.14, respectively). Strain rate and FAC were positively correlated with heart rate (r 2 ¼ 0.35 and 0.31, respectively). Allometric scaling generated bodyweightbased reference intervals for these RV systolic function indices. Conclusions: Echocardiographic indices of RV systolic function are feasible to obtain, repeatable, and affected by bodyweight. Studies of these indices in dogs with cardiovascular disease are needed. ª 2014 Elsevier B.V. All rights reserved.

Abbreviations 2D CV FAC RV S’

two-dimensional coefficient of variation fractional area change right ventricular pulsed wave tissue Doppler imagingderived systolic myocardial velocity of the lateral tricuspid annulus SD standard deviation STE speckle tracking echocardiography TAPSE tricuspid annular plane systolic excursion TDI tissue Doppler imaging

Introduction The right ventricle is affected by a number of diseases, including pulmonary hypertension caused by lung disease, pulmonary vascular disease, or left-sided heart disease; right ventricular (RV) cardiomyopathies, such as arrhythmogenic RV cardiomyopathy; pericardial disease; pulmonary or tricuspid valve malformations; cardiac shunts; and complex congenital heart disease. The clinical recognition of RV dysfunction in veterinary medicine is underdeveloped and has traditionally relied on qualitative assessment or overt signs of rightsided congestive heart failure. The qualitative assessment of RV structure and function in human patients is inaccurate, with low interobserver agreement.1 Consequently, measured and calculated indices that quantify RV function might be clinically useful in identifying the presence and progression of RV dysfunction. The importance of the quantitative assessment of RV function is increasingly apparent in people affected with both cardiac and non-cardiac diseases.2 Quantitative analysis of RV function provides prognostic data and guides the clinical decision-making process not only in right heartspecific diseases3 but also left heart disorders,

including mitral and aortic valve disease4e6 and dilated cardiomyopathy,7e12 often independent of pulmonary hypertension status. Similar studies of quantitative RV function in dogs could potentially be of similar clinical value. However, when compared to the left ventricle, the assessment of RV function is more difficult owing to its complex geometry. Specific anatomical challenges include separate inflow and outflow regions, prominent endocardial trabeculations, ventricular interdependence, and the marked load-dependence of most indices of RV function.13 Echocardiography is the most practical method for assessment of RV structure and function in veterinary medicine as it is noninvasive, readily available, relatively inexpensive, and does not require general anesthesia. Both guidelines and reference intervals are available for a number of RV echocardiographic indices in people.14 Although each index has inherent advantages and disadvantages, nearly all human RV indices have been validated against a catheterization- or magnetic resonance imaging derived gold standard. These include the M-mode-derived tricuspid annular plane systolic excursion (TAPSE), the 2dimensional (2D) correlate to RV ejection fraction e percent fractional area change (FAC), tissue Doppler imaging (TDI)-derived systolic myocardial velocity of the lateral tricuspid annulus (S’), and speckle-tracking echocardiography (STE)-derived strain and strain rate.14e18 Aside from TAPSE,19 canine reference intervals for RV systolic function indices based on estimates of central tendencies in a large sample of the healthy canine population are lacking. Such reference intervals along with repeatability data are essential prior to widespread clinical application of echocardiographic indices in diseased dogs. As several echocardiographic indices of cardiac structure and function are known to be affected by age, gender and body size in humans20e27 and in animals,28e32 these variables also should be considered when establishing reference intervals. The impact of bodyweight particularly warrants

Echocardiographic reference intervals for RV function in dogs consideration for the most accurate assessment of reference intervals given the wide range of body sizes in dogs. The aforementioned considerations led us to evaluate 5 different, but complementary echocardiographic indices of RV systolic function. The first study objective was determination of the feasibility, repeatability and intra- and interobserver variability of TAPSE, FAC, S’, and STEderived global longitudinal RV free wall strain and strain rate. The second objective was to explore the statistical effects of age, gender, heart rate, and bodyweight on those indices. Finally, based on these data, clinically-applicable reference intervals were generated.

Animals, materials and methods All procedures in this study were approved by the Institutional Animal Care and Use Committee and the Veterinary Medical Center Clinical Research and Teaching Advisory Committee at The Ohio State University. Written consent authorizing participation of dogs in the study was obtained from all dog owners.

Animals A convenience sample of 80 privately-owned healthy, mature dogs 8 months of age and of varying breed and bodyweight (n ¼ 40 > 15 kg; n ¼ 40
15 kg) were recruited for this study from members of The Ohio State University College of Veterinary Medicine. Dogs were determined to be healthy and without cardiac or respiratory diseases based on medical history, routine physical examination, cardiovascular examination, and a thorough screening echocardiogram. Exclusion criteria for the study were: 1) pathologic heart murmur, gallop sound, or (non-sinus) arrhythmia; 2) history of respiratory disease; 3) taking medications known to affect the cardiovascular or respiratory systems; 4) uncooperative temperament that might require sedation for an echocardiogram; 5) Boxer dogs and English bulldogs (due to risk of occult arrhythmogenic RV cardiomyopathy); and 6) cardiac abnormalities identified on a baseline 2D, M-mode, and Doppler echocardiographic study. Right heart valve regurgitation evident on color Doppler echocardiography was defined as physiologic if silent to auscultation and associated with normal valve morphology. Dogs with physiologic tricuspid or pulmonary valve regurgitation were not excluded due to the high prevalence of this finding in healthy dogs.33

85

Echocardiographic examination Conventional and Doppler echocardiography All echocardiographic studies were performed by the same investigator (LCV) using a GE Vivid 7 echocardiographic systema with transducer selection (4, 7, or 10 MHz nominal frequency) matched to the size of the dog and preset for optimal imaging. Echocardiographic recordings were made with a simultaneous ECG, and all raw data were captured digitally for off-line analysis at a digital workstation. Standard imaging planes34 were utilized with the dogs manually restrained in right and left lateral recumbency without the use of sedation. Echocardiographic indices of RV systolic function All the indices of RV systolic function were acquired from the left apical 4-chamber view optimized for the right heart. This involved transducer placement 1 intercostal space cranial to the standard left apical 4-chamber view with varying degrees of caudal angulation. Care was taken to maximize the RV longitudinal dimension and to exclude the left ventricular outflow tract to avoid foreshortening of the RV. The ultrasound system was adjusted to optimize RV myocardial and endomyocardial border resolution. The examiner attempted to record images during periods of quiet/calm respiration. At least 10 cardiac cycles of each RV function index were acquired and stored for off-line analysis. Measured images were chosen based on technical adequacy, without regard to the respiratory cycle as translational motion with ventilation precluded using consecutive cardiac cycles. The value recorded for each RV function index at each time point was determined from an average of 5 representative cardiac cycles. The heart rate value recorded represented the average heart rate of each of the 5 cardiac cycles used to determine the RV function index value. The TAPSE measurement consisted of quantifying the maximal longitudinal displacement of the lateral tricuspid valve annulus toward the RV apex during systole and was generated from M-mode recordings with the cursor as parallel as possible to the majority of the RV free wall (Fig. 1). In order to avoid underestimating TAPSE values, the anatomic M-mode technique35 was activated infrequently (<10% of recordings) on stored 2D cine loops. Anatomic M-mode was used when conventional Mmode cursor alignment to the RV free wall was a Vivid 7 Dimension with EchoPac software package, version BT09, GE Medical Systems, Waukesha, WI, USA.

86

L.C. Visser et al.

Figure 1 Representative measurement of tricuspid annular plane systolic excursion (TAPSE). This measurement quantifies maximal longitudinal displacement of the lateral tricuspid valve annulus toward the right ventricular (RV) apex during systole using M-mode echocardiography. Note the cursor (white dotted line) is aligned as parallel as possible to the majority of the RV free wall. RA, right atrium; RV, right ventricle.

suboptimal, or when tricuspid annular motion on the M-mode recording was unclear. The TAPSE measurements were made at sweep speeds of at least 66 mm/s. Measurements of RV area for FAC determination were obtained by tracing the RV endocardial border at end-diastole (RVAD) and end-systole (RVAS) as shown in Figure 2. Right ventricular percent FAC was calculated using the formula: FAC ¼ ([RVAD  RVAS]/RVAD)  100.

Pulsed-wave TDI velocities of longitudinal myocardial motion at the lateral tricuspid annulus were obtained to measure peak systolic annular velocity (S’) as shown in Figure 3. For accurate TDI imaging, the cursor was aligned as parallel as possible to the longitudinal plane of the RV free wall, and recording frame rate was at least 125 frames/s. Measurements of S’ were made at sweep speeds of at least 66 mm/s with sample volumes ranging from 1 to 4 mm.

Figure 2 Representative measurement of right ventricular (RV) percent fractional area change (FAC) from a 2D echocardiographic image. Measurements of RV area were obtained by tracing the RV endocardial border (dotted lines) at end-diastole (RVAD) and end-systole (RVAS). The FAC is calculated using the formula: FAC ¼ ((RVAD  RVAS)/ RVAD)  100. RA, right atrium; RV, right ventricle.

Echocardiographic reference intervals for RV function in dogs

87

Figure 3 Representative measurement of pulsed-wave tissue Doppler imaging (TDI)-derived peak systolic longitudinal myocardial motion velocity at the lateral tricuspid annulus (S’). RA, right atrium; RV, right ventricle.

Strain and strain rate measurements were calculated with proprietary 2D speckle-tracking softwareb using the left ventricular 2-chamber algorithm, as no defined RV STE algorithm was available at the time of this study. Because the optimal frame rate for canine RV STE is currently unknown, 9 2D cine loops each were acquired as follows: 3 loops at maximal, 3 loops at 1 less than maximal, and 3 loops at 2 less than maximal frame rates. Event timing for pulmonary valve opening and closure relative to the ECG was determined individually for each study by measuring a continuous wave Doppler recording of RV outflow ejection velocity. Only RV free wall longitudinal strain and strain rate were analyzed as longitudinal motion has been recognized to be the major contributor to RV contraction in the dog.36 The region of interest for STE was defined by manually tracing the RV free wall endocardial border from the level of the tricuspid valve annulus to the RV apex with manual adjustments to incorporate the entire RV free wall myocardial thickness. Individual RV segments were then visually analyzed to assure adequate myocardial tracking by the software and manually adjusted if necessary. In general, tracking was accepted if both visual inspection and software inspection (green color coding) confirmed it was adequate (Fig. 4). However, on rare occasion, when software approval was unobtainable it b EchoPAC 2D Strain software, Q-Analysis (strain module), version 6.1, GE Medical Systems, Waukesha, WI, USA.

was manually overridden so long as visual inspection of myocardial tracking was considered appropriate. The frame rate that provided the most accurate tissue tracking based on visual and software inspection was chosen for STE-derived strain and strain rate (generally >80 frames/s). Strain and strain rate values were generated by the software for each of 3 myocardial segments (basilar, mid, and apical myocardium of the RV free wall) in addition to the global strain and strain rate from the entire RV (considered as a single segment and not a mean of the 3 segments). Only global longitudinal systolic strain and strain rate values of the RV free wall were used in this study and were determined as the maximal (most negative) systolic point on the respective global strain or strain rate curve prior to pulmonary valve closure (Figs. 5 and 6).

Repeatability, intra- and interobserver measurement variability Day-to-day repeatability (recording variability or reproducibility) of each RV function index was evaluated by having each dog undergo 2 echocardiographic studies performed between 3 and 20 days apart. Intraobserver measurement variability was determined by having the same individual (LCV) measure RV function indices from 6 randomly selected echocardiographic studies on 3 separate occasions. Interobserver measurement variability was determined by having 2 trained

88

L.C. Visser et al.

Figure 4 Representative 2D echocardiographic image of the right heart from the left apical four-chamber view optimized for the right ventricle in which longitudinal segments have been designated by the proprietary software as basilar inferior (basInf; yellow), mid inferior (midInf; light blue) and apical inferior (apInf; green) due to the left ventricular 2Ch algorithm (as there was no right ventricular speckle-tracking echocardiography algorithm at the time of study). The green colored bars with a “V” below the labeled segments indicate that tissue tracking by the software is adequate. RA, right atrium; RV, right ventricle.

individuals (LCV & BAS), blinded to each other’s measurements, measure all RV function indices from 6 randomly selected echocardiographic studies. The 6 selected echocardiographic studies used to determine the intra- and interobserver measurement variability were determined by randomlyc choosing 3 echocardiographic studies from dogs weighing
15 kg and 3 echocardiographic studies from dogs weighing >15 kg. Anatomic Mmode for TAPSE determination was not used in any of the 6 studies of measurement variability data, nor was a manual override for STE-derived strain and strain rate employed for these measurements.

Statistical analysis All statistical analyses were performed using commercial software packages.d,e For the purpose of the reference intervals, values for each RV function index and heart rate were generated as c

http://www.randomizer.org/. IBM SPSS Statistics, version 21, IBM Corp, Armonk, NY, USA. e MedCalc, version 12.7.4, MedCalc Software, Ostend, Belgium. d

the average of the 2 day-to-day repeatability data sets (pooled data) from each dog. Descriptive statistics (mean, median, standard deviation [SD], and 95% confidence intervals or interquartile range) were calculated for all RV function indices. Normality testing for continuous data consisted of visual inspection of the probability plots and the D’Agostino-Pearson test. A value of P < 0.05 was considered statistically significant. For all linear regression models, assumptions of linearity, homoscedasticity, and normality of the residuals were evaluated by inspection of the standardized residual plots and probability plots. Normality of the standardized residuals was also assessed with the D’Agostino-Pearson test. Standardized residual plots and Cook’s distances were used to identify possible outliers and influential data points on the model and if Cook’s distances of greater than 1.0 were encountered, the data point was excluded from further analysis.37 Multiple linear regression analysis was used to explore the relationship between the indices of RV systolic function and bodyweight (in kg), age (in months), and heart rate (in beats/min). An unpaired t-test (or ManneWhitney U test) was used to test for

Echocardiographic reference intervals for RV function in dogs

89

Figure 5 Representative snapshot of the workstation output for right ventricular (RV) free wall longitudinal strain. A reference 2D image of the right heart can be seen in the upper left, a color map is shown in the lower left to display the change in strain over one cardiac cycle (red ¼ negative; blue ¼ positive), and the remainder of the snapshot includes 3 regional (basilar ¼ yellow, mid ¼ light blue, apical ¼ green) and global (dotted line) strain curves over time in relation to the ECG (bottom). Only systolic global longitudinal strain of the RV free wall prior to pulmonary valve closure was utilized in this study (arrow). The “AVC” corresponds to pulmonary valve closure (instead of aortic valve closure [AVC]) determined from a continuous wave Doppler recording of pulmonary outflow timed with the ECG. RA, right atrium; RV, right ventricle.

differences between male and female dogs for each RV function index. To account for differences that can be expected with the large variation in bodyweight of the dogs, weight-dependent regression-based reference intervals were determined. RV function indices were scaled to bodyweight and several regression models were tested for fit, including linear, second-order polynomial, third-order polynomial and allometric (power) models. For the purpose of this study, the simplest mathematical model (i.e., the model with the fewest number of predictors) achieving the highest degree of statistical significance (i.e., the largest F-statistic) and adjusted r2 value was considered the best model of fit. Constants for allometric modeling were derived using the logarithmic form of the allometric scaling equation: log (Y) ¼ log (a) þ b  log (M), where a is the proportionality constant, b is the scaling exponent, Y represents the RV function index, and M represents bodyweight.38 Simple linear regression analysis yields the constant b, which is the slope of the regression line and the constant a, which is the antilog (log1) of the y-intercept of the regression line. The 95% prediction intervals for the linear regression line of

best fit were then used to calculate the recommended lower and upper reference intervals. The average percent coefficient of variation (CV) was used to quantify day-to-day repeatability for the 2 time points, intraobserver measurement variability and interobserver measurement variability, where percent CV ¼ (SD of the measurements/average of measurements)  100.

Results Animals The study sample consisted of 80 dogs with a median age of 4.1 years (minimum ¼ 0.66 years; maximum ¼ 9 years; interquartile range ¼ 2.2e6.4 years) and a median bodyweight of 15.8 kg (minimum ¼ 3.9 kg; maximum ¼ 42.3 kg; interquartile range ¼ 8.2e27.2 kg). The mean heart rate during the studies was 106  22 beats/min (minimum ¼ 57 beats/min; maximum ¼ 158 beats/ min). Tricuspid regurgitation was observed by color Doppler imaging in 23 of 80 dogs (29%) and graded as mild in all cases. Thirty-six dogs were

90

L.C. Visser et al.

Figure 6 Representative snapshot of the workstation output for right ventricular (RV) free wall longitudinal strain rate. A reference 2D image of the right heart can again be seen in the upper left, a color map is shown in the lower left to display the change in strain rate over one cardiac cycle (red ¼ negative; blue ¼ positive), and the remainder of the snapshot includes 3 regional (basilar ¼ yellow, mid ¼ light blue, apical ¼ green) and global (dotted line) strain rate curves over time in relation to the ECG (bottom). Only systolic global longitudinal strain rate of the RV free wall was utilized in this study (arrow). The “AVC” corresponds to pulmonary valve closure (instead of aortic valve closure [AVC]. RA, right atrium; RV, right ventricle.

castrated males and 44 were spayed females. Forty-one dogs were mixed breeds, 5 were Pugs, 4 dogs each were Boston Terriers, Labrador Retrievers, Golden Retrievers, and Miniature Schnauzers, 2 dogs each were Cavalier King Charles Spaniels, Rat Terriers, Italian Greyhounds, Chihuahuas, Beagles, and German Shepherds. The other breeds (Border Collie, Wheaton Terrier, Bloodhound, Miniature Pinscher, Greyhound, Pembroke Welsh Corgi, English Setter, Toy Poodle, Pomeranian, and Papillon) were each represented once.

Table 1

Repeatability, intra- and interobserver measurement variability Each RV function index could be obtained in all dogs. Descriptive statistics for each of the RV function indices from the echocardiographic studies are summarized in Table 1. Average CV for repeatability, intra- and interobserver measurement variability of all RV function indices were considered low (<10%) and are summarized in Table 2.

Descriptive statistics for the right ventricular systolic function indices in 80 conscious healthy dogs.

RV function index TAPSE (mm) FAC (%) S’ (cm/s) Strain  1 (%) Strain Rate  1 (s1)

Mean 13.75 46.50 13.42 28.62 3.29

Median 13.37 46.90 13.22 28.58 3.20

SD 3.41 6.57 3.97 4.02 0.91

95% CIa/IQRb b

11.40e15.53 45.04e47.96a 12.54e14.31a 27.73e29.52a 3.09e3.49a

MineMax 8.53e25.00 32.83e62.25 6.83e26.13 19.95e46.41 1.56e6.59

RV, right ventricular; TAPSE, tricuspid annular plane systolic excursion; FAC, fractional area change; S’, pulsed wave tissue Doppler-derived lateral tricuspid annular longitudinal peak systolic velocity; SD, standard deviation; CI, confidence interval of the mean; IQR, interquartile range. a Values represent the 95% confidence interval of the mean. b Values represent the interquartile range (data set was not normally-distributed).

Echocardiographic reference intervals for RV function in dogs

91

Table 2 Repeatability, intra- and interobserver measurement variability of echocardiographic indices of right ventricular systolic function in conscious healthy dogs. RV function index

Average coefficient of variation (%) Repeatability (n ¼ 80)

Intra-observer variability (n ¼ 6)

Inter-observer variability (n ¼ 6)

4.7 4.3 7.8 4.0 8.5

3.3 4.9 4.0 2.9 7.3

3.6 6.4 7.9 8.4 8.6

TAPSE FAC S’ Strain Strain Rate See Table 1 for key.

RV function indices in conscious healthy dogs Each RV function index was significantly correlated (all P
0.023) to bodyweight. Only FAC and strain rate exhibited a significant moderate positive correlation (P
0.001) with heart rate. Age did not significantly contribute to the model predicting any RV function index. The zero-order (univariate) and part (semi-partial) correlation coefficients for the independent variables are presented in Table 3. Part correlations indicate the unique variation shared between a predictor and the dependent variable while controlling for the effects of other predictors. There was no statistically significant difference between male and female dogs for any RV function index. Regression using allometric (power) scaling exhibited the best model fit compared to simple

linear, second-order polynomial and third-order polynomial regression models for all RV function indices and bodyweight. All RV function indices were significantly correlated to bodyweight (P
0.001) with variable coefficients of determination (Table 4). Based on allometric scaling and data from these 80 dogs, reference intervals for each RV function index across bodyweights were estimated (Table 5). These represent the 95% prediction intervals for the best fit regression line derived from the allometric equations.

Discussion The results of this study show that TAPSE, FAC, S’, strain, and strain rate are feasible to acquire in dogs of varying bodyweight. These variables are reproducible and demonstrate clinically

Table 3 Correlation coefficients generated from multiple linear regression analyses demonstrating the relationship between indices of right ventricular systolic function and bodyweight, age, and heart rate. RV function index TAPSE

FAC

S’

Strain

Strain rate

Predictors

Zero-order correlation

Part correlation

P-value

Bodyweight Heart rate Age Bodyweight Heart rate Age Bodyweight Heart rate Age Bodyweight Heart rate Age Bodyweight Heart rate Age

0.858 0.385 0.149 0.410 0.497 0.059 0.550 0.097 0.037 0.339 0.215 0.136 0.610 0.554 0.020

0.755 0.036 0.030 0.231 0.345 0.070 0.555 0.126 0.056 0.249 0.083 0.099 0.403 0.308 0.025

<0.001a 0.544 0.616 0.018a 0.001a 0.470 <0.001a 0.188 0.555 0.023a 0.441 0.356 <0.001a <0.001a 0.763

Zero-order correlation coefficients represent the variation between the predictor and the RV function index while ignoring the influence of the other predictors. Part correlations represent the unique variation each predictor shares with the RV function index while controlling (removing) influence of the other predictors. See Table 1 for key. a Denotes statistical significance.

92

L.C. Visser et al.

Table 4 Results of simple linear regression on logarithmically transformed right ventricular function indices and bodyweight. RV function index TAPSE FAC S’ Strain Strain rate

Log (a)

a

95% prediction interval for a

b

Std error of Y est

r2

0.781 1.776 0.839 1.540 0.808

6.039 59.70 6.898 34.67 6.427

4.772e7.644 46.34e76.92 4.262e11.18 26.85e44.78 4.289e9.629

0.297 0.097 0.233 0.075 0.263

0.0514 0.0553 0.1052 0.0558 0.0882

0.749 0.216 0.306 0.140 0.440

Linear regression analysis of the logarithmic form of the allometric equation: log (Y) ¼ log (a) þ b  log (M), where a is the proportionality constant, b is the scaling exponent, Y represents the RV function index, and M represents bodyweight, yields the constant b, which is the slope of the regression line and the constant a, which is the antilog (log1) of the y-intercept of the regression line. The equation can be rewritten as the allometric equation: Y ¼ a  Mb. Std error of Y est, standard error of the Y estimate. See Table 1 for the remainder of the key. All correlations were statistically significant (P
0.001).

acceptable intra- and interobserver measurement variability. Importantly, each of these 5 indices are significantly related to bodyweight with TAPSE exhibiting a strong positive correlation, S’ and strain rate moderate positive and negative correlations, respectively, and FAC and strain exhibiting weak negative correlations to bodyweight. Moderate positive correlations between strain rate and FAC to heart rate were also found, which might have importance clinically when heart rates deviate significantly from normal values. Last, bodyweight-specific reference intervals from a relatively large sample of conscious healthy dogs are provided so that a quantitative assessment of RV systolic function can be performed.

Despite encouragement,39 the echocardiographic assessment of RV structure and function has thus far received little attention in dogs. Previous studies on the assessment of RV systolic function have focused on comparing dogs with natural or induced disease to a relatively small number of controls40e44 or reported combined or comparative RV and left ventricular systolic function indices.45,46 While these studies have provided valuable background information, it is difficult to advance these results to wider clinical use without first establishing intervals that potentially differentiate normal from diseased dogs. Other investigators have examined echocardiographic indices of RV function in relatively large

Table 5 Bodyweight-dependent reference intervals (95% prediction intervals) of 5 echocardiographic indices of right ventricular systolic function assessed in 80 conscious healthy dogs. Bodyweight (kg) 3 4 5 7 9 12 15 20 25 30 35 40 45

RV function index TAPSE (mm) 6.6e10.6 7.2e11.5 7.7e12.3 8.5e13.6 9.2e14.7 10.0e16.0 10.7e17.1 11.6e18.6 12.4e19.9 13.1e21.0 13.7e22.0 14.3e22.8 14.8e23.7

See Table 1 for key. a allometric equation b allometric equation c allometric equation d allometric equation e allometric equation

with with with with with

95% 95% 95% 95% 95%

a

FAC (%)

b

41.7e69.1 40.5e67.2 39.6e65.8 38.4e63.7 37.4e62.2 36.4e60.5 35.6e59.2 34.7e57.5 33.9e56.3 33.3e55.3 32.8e54.5 32.4e53.8 32.0e53.2 prediction prediction prediction prediction prediction

intervals: intervals: intervals: intervals: intervals:

Y Y Y Y Y

¼ ¼ ¼ ¼ ¼

S’ (cm/s)c

Strain  1 (%)d

Strain rate  1 (s1)e

5.5e14.4 5.9e15.4 6.2e16.3 6.7e17.6 7.1e18.7 7.6e19.9 8.0e21.0 8.6e22.5 9.0e23.7 9.4e27.4 9.8e25.6 10.1e26.4 10.3e27.1

24.7e41.2 24.2e40.4 23.8e39.7 23.2e38.7 22.8e38.0 22.3e37.2 21.9e36.5 21.4e35.8 21.1e35.2 20.8e34.7 20.6e34.3 20.4e34.0 20.2e33.7

3.2e7.2 3.0e6.7 2.8e6.3 2.6e5.8 2.4e5.4 2.2e5.0 2.1e4.7 2.0e4.4 1.8e4.1 1.8e3.9 1.7e3.8 1.6e3.6 1.6e3.5

4.777 to 7.640  M0.297. 46.34 to 76.92  M0.097. 4.262 to 11.178  M0.233. 26.851 to 44.776  M0.075. 4.289 to 9.629  M0.263.

Echocardiographic reference intervals for RV function in dogs samples of healthy dogs. These studies have evaluated color TDI-derived RV myocardial velocities,46 TAPSE,19 RV systolic time intervals,41,45 RV shortening fraction,41 and a RV myocardial (Tei) index.45 However, with the exception of TAPSE, reference intervals were either not reported or were reported as central tendencies for estimating the mean (e.g., confidence intervals of the mean or SD) and not as central tendencies for predicting values for the canine population. As shown in the present study, others have documented a significant effect of bodyweight on RV function indices.19,45,46 This finding highlights the need to consider bodyweight and the need for bodyweightspecific reference intervals and calculations predicting central tendencies of the population, as performed in the current investigation. A similar approach was undertaken by Pariaut and colleagues19 who examined TAPSE in 50 normal dogs in addition to dogs affected with pulmonary hypertension. While using a slightly different analysis, they also documented a strong (r2 ¼ 0.83) correlation with bodyweight and reported weightspecific reference values comparable to those reported herein. The generation of reference intervals not only helps to distinguish healthy from diseased dogs, but these cutoffs could influence clinical decisions regarding diagnosis, therapy and prognosis. Based on current guidelines for the echocardiographic assessment of the right heart in adult humans,14 RV function indices, including TAPSE, FAC, and S’, are presented as single cutoff values. This methodology is problematic in dogs given the wide range of body sizes, a problem shared with pediatric cardiology. Recommendations for the use of body surface area and conversion of indices to Z-scores have been suggested in current guidelines for growing children.47 However, despite these recommendations, more recent studies48,49 show that standardization of measurements is still a problem. In the current study, the method utilized to establish reference intervals involved regression analysis, for which allometric (power) scaling provided the best fit for the function indices studied. Allometry describes the disproportionate changes in shape, size, or function observed when comparing separate isolated features in animals that vary in body size.38 To quantify this relationship the size of an organ or its function may be expressed as a power function of body mass or weight. Similar to the present study, several investigators have used allometric scaling to normalize indices of mainly left ventricular structure

93

and function to bodyweight in laboratory animals and dogs.50e54 A recent study in people compared the qualitative and quantitative assessments of RV function to quantitative measures using cardiac magnetic resonance imaging as a gold standard. In this study TAPSE, FAC, S’, and a RV index of myocardial performance (Tei index) were superior to qualitative assessment with increased accuracy and improved interobserver agreement.1 This highlights the limitations of subjective RV function analysis. Our study also showed that, with the addition of strain and strain rate, the same RV function indices (TAPSE, FAC and S’) exhibited acceptable inter- and intraobserver agreement in normal dogs. This was also demonstrated in a recent study19 showing that TAPSE exhibits tolerable inter- and intraobserver variability. Furthermore, the current study demonstrated that in conscious healthy dogs TAPSE, FAC, S’ (pulsed wave-derived), strain, and strain rate were highly reproducible. This is comparable to studies in people,14,55 but, to the authors’ knowledge, reproducibility data for the studied RV function indices in dogs have not been reported. One limitation of the current study is the lack of longitudinal follow-up of the dogs used to define the reference intervals. Without follow-up, it cannot be certain that the studied dogs were free of subclinical cardiac disease that could affect RV function thereby skewing the reference intervals. Although respiratory variation and cardiac cycle length can affect RV size and function, we did not attempt to control for respiratory cycle or sinus arrhythmia, instead averaging 5 different cardiac cycles per data point. Respiration also leads to significant translational motion that can alter image planes and optimal alignment to contraction. Another limitation is the sample size of 80 dogs, particularly when compared to sample sizes used for generation of reference intervals in human medicine. However when compared to other veterinary echocardiographic studies of reference intervals,56 the current report represents one of the larger published samples. From a statistical perspective, at least 120 observations are recommended for generation of reference intervals in order to provide the most reliable cutoffs.57 To this end, RV function index values that are very close to the recommended cutoffs should be interpreted with caution. Caution is also warranted when applying the reference values determined in this study to dogs outside the bodyweights encountered in the current study (<3.9 kg and >42.3 kg).

94

L.C. Visser et al.

Conclusions The current study demonstrates that 5 indices of RV systolic function (TAPSE, FAC, S’, strain, and strain rate) are feasible to acquire, are reproducible, and had intra- and interobserver measurement variability within clinically acceptable limits. Bodyweight should be considered when interpreting these RV function indices. The dog’s heart rate should be considered when interpreting FAC and strain rate values. The current study provides bodyweight-specific reference intervals for 5 indices of RV function thereby encouraging the quantitative assessment of RV systolic function in dogs with cardiovascular disease.

6.

7.

8.

9.

10.

Conflict of interest statement The authors declare no conflict of interest. 11.

Acknowledgments This study was supported by the American Kennel Club Canine Health Foundation Clinician-Scientist Fellowship Program (LCV). The authors would like to thank Tammy Muse, Patti Mueller, and Drs. Kay Drake and Kyla Morgan for technical assistance and the dog owners of the OSU CVM community for volunteering their dogs.

References 1. Ling LF, Obuchowski NA, Rodriguez L, Popovic Z, Kwon D, Marwick TH. Accuracy and interobserver concordance of echocardiographic assessment of right ventricular size and systolic function: a quality control exercise. J Am Soc Echocardiogr 2012;25:709e713. 2. Voelkel NF, Quaife RA, Leinwand LA, Barst RJ, McGoon MD, Meldrum DR, Dupuis J, Long CS, Rubin LJ, Smart FW, Suzuki YJ, Gladwin M, Denholm EM, Gail DB. Right ventricular function and failure: report of a national heart, lung, and blood institute working group on cellular and molecular mechanisms of right heart failure. Circulation 2006;114:1883e1891. 3. Haddad F, Doyle R, Murphy DJ, Hunt SA. Right ventricular function in cardiovascular disease, part II: pathophysiology, clinical importance, and management of right ventricular failure. Circulation 2008;117:1717e1731. 4. Le Tourneau T, Deswarte G, Lamblin N, Foucher-Hossein C, Fayad G, Richardson M, Polge AS, Vannesson C, Topilsky Y, Juthier F, Trochu JN, Enriquez-Sarano M, Bauters C. Right ventricular systolic function in organic mitral regurgitation: impact of biventricular impairment. Circulation 2013;127: 1597e1608. 5. Hochreiter C, Niles N, Devereux RB, Klingfield P, Borer JS. Mitral regurgitation: relationship of noninvasive descriptors of right and left ventricular performance to clinical and

12.

13.

14.

15.

16.

17.

18. 19.

20.

hemodynamic findings and to prognosis in medically and surgically treated patients. Circulation 1986;73:900e912. Nagel E, Stuber M, Hess OM. Importance of the right ventricle in valvular heart disease. Eur Heart J 1996;17: 829e836. Meluzin J, Spinarova L, Hude P, Krejci J, Kinci V, Panovsky R, Dusek L. Prognostic importance of various echocardiographic right ventricular functional parameters in patients with symptomatic heart failure. J Am Soc Echocardiogr 2005;18:435e444. Ghio S, Gavazzi A, Campana C, Inserra C, Klersy C, Sebastiani R, Arbustini E, Recusani F, Tavazzi L. Independent and additive prognostic value of right ventricular systolic function and pulmonary artery pressure in patients with chronic heart failure. J Am Coll Cardiol 2001;37:183e188. Di Salvo TG, Mathier M, Semigran MJ, Dec GW. Preserved right ventricular ejection fraction predicts exercise capacity and survival in advanced heart failure. J Am Coll Cardiol 1995;25:1143e1153. La Vecchia L, Bedogni F, Castellani A, Martini M, Paccanaro M, Sartori M, Bozzola L, Bevilacqua P, Vincenzi M. Assessment of right ventricular function and interstitial fibrosis in idiopathic dilated cardiomyopathy: hemodynamic correlates and prognostic value. G Ital Cardiol 1998;28: 513e523. Juilliere Y, Barbier G, Feldmann L, Grentzinger A, Danchin N, Cherrier F. Additional predictive value of both left and right ventricular ejection fractions on long-term survival in idiopathic dilated cardiomyopathy. Eur Heart J 1997;18:276e280. de Groote P, Millaire A, Foucher-Hossein C, Nuque O, Marchandise X, Ducloux G, Lablanche JM. Right ventricular ejection fraction is an independent predictor of survival in patients with moderate heart failure. J Am Coll Cardiol 1998;32:948e954. Haddad F, Hunt SA, Rosenthal DN, Murphy DJ. Right ventricular function in cardiovascular disease, part I: anatomy, physiology, aging, and functional assessment of the right ventricle. Circulation 2008;117:1436e1448. Rudski LG, Lai WW, Afilalo J, Hua L, Handschumacher MD, Chandrasekaran K, Soloman SD, Louie EK, Schiller NB. Guidelines for the echocardiographic assessment of the right heart in adults: a report from the American Society of Echocardiography endorsed by the European Association of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. J Am Soc Echocardiogr 2010;23:685e713. Horton KD, Meece RW, Hill JC. Assessment of the right ventricle by echocardiography: a primer for cardiac sonographers. J Am Soc Echocardiogr 2009;22:776e792. Mertens LL, Friedberg MK. Imaging the right ventricleecurrent state of the art. Nat Rev Cardiol 2010;7: 551e563. Jurcut R, Giusca S, La Gerche A, Vasile S, Ginghina C, Voigt JU. The echocardiographic assessment of the right ventricle: what to do in 2010? Eur J Echocardiogr 2010;11: 81e96. Sheehan F, Redington A. The right ventricle: anatomy, physiology and clinical imaging. Heart 2008;94:1510e1515. Pariaut R, Saelinger C, Strickland KN, Beaufrere H, Reynolds CA, Vila J. Tricuspid annular plane systolic excursion (TAPSE) in dogs: reference values and impact of pulmonary hypertension. J Vet Intern Med 2012;26: 1148e1154. Koestenberger M, Ravekes W, Everett AD, Stueger HP, Heinzi B, Gamillscheg A, Cvirn G, Boysen A, Fandi A, Nagel B. Right ventricular function in infants, children and

Echocardiographic reference intervals for RV function in dogs

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

adolescents: reference values of the tricuspid annular plane systolic excursion (TAPSE) in 640 healthy patients and calculation of z score values. J Am Soc Echocardiogr 2009;22: 715e719. Koestenberger M, Nagel B, Ravekes W, Avian A, Heinzl B, Cvirn G, Fritsch P, Fandl A, Rehak T, Gamillscheg A. Reference values of tricuspid annular peak systolic velocity in healthy pediatric patients, calculation of z score, and comparison to tricuspid annular plane systolic excursion. Am J Cardiol 2012;109:116e121. Pfaffenberger S, Bartko P, Graf A, Pernicka E, Babayev J, Lolic E, Bonderman D, Baumgartner H, Maurer G, Mascherbauer J. Size matters! impact of age, sex, height, and weight on the normal heart size. Circ Cardiovasc Imaging 2013;6:1073e1079. D’Oronzio U, Senn O, Biaggi P, Gruner C, Jenni R, Tanner FC, Greutmann M. Right heart assessment by echocardiography: gender and body size matters. J Am Soc Echocardiogr 2012; 25:1251e1258. Willis J, Augustine D, Shah R, Stevens C, Easaw J. Right ventricular normal measurements: time to index? J Am Soc Echocardiogr 2012;25:1259e1267. Eidem BW, McMahon CJ, Cohen RR, Wu J, Finkelshteyn I, Kovalchin JP, Ayres NA, Bezold LI, O’Brian Smith E, Pignatelli RH. Impact of cardiac growth on Doppler tissue imaging velocities: a study in healthy children. J Am Soc Echocardiogr 2004;17:212e221. Kou S, Caballero L, Dulgheru R, Voilliot D, De Sousa C, Kacharava G, Athanassopoulos GD, Barone D, Baroni M, Cardim N, Gomez De Diego JJ, Hagendorff A, Henri C, Hristova K, Lopez T, Magne J, De La Morena G, Popescu BA, Penicka M, Ozyigit T, Rodrigo Carbonero JD, Salustri A, Van De Veire N, Von Bardeleben RS, Vinereanu D, Voigt JU, Zamorano JL, Donal E, Lang RM, Badano LP, Lancellotti P. Echocardiographic reference ranges for normal cardiac chamber size: results from the NORRE study. Eur Heart J Cardiovasc Imaging 2014;15:680e690. Maffessanti F, Muraru D, Esposito R, Gripari P, Ermacora D, Santoro C, Tamborini G, Galderisi M, Pepi M, Badano LP. Age-, body size-, and sex-specific reference values for right ventricular volumes and ejection fraction by threedimensional echocardiography: a multicenter echocardiographic study in 507 healthy volunteers. Circ Cardiovasc Imaging 2013;6:700e710. Jacobs G, Mahjoob K. Multiple regression analysis, using body size and cardiac cycle length, in predicting echocardiographic variables in dogs. Am J Vet Res 1988;49: 1290e1294. Morrison SA, Moise NS, Scarlett J, Mohammed H, Yeager AE. Effect of breed and body weight on echocardiographic values in four breeds of dogs of differing somatotype. J Vet Intern Med 1992;6:220e224. Goncalves AC, Orton EC, Boon JA, Salman MD. Linear, logarithmic, and polynomial models of M-mode echocardiographic measurements in dogs. Am J Vet Res 2002;63: 994e999. Brown DJ, Rush JE, MacGregor J, Ross Jr JN, Brewer B, Rand WM. M-mode echocardiographic ratio indices in normal dogs, cats, and horses: a novel quantitative method. J Vet Intern Med 2003;17:653e662. Sisson D, Schaeffer D. Changes in linear dimensions of the heart, relative to body weight, as measured by M-mode echocardiography in growing dogs. Am J Vet Res 1991;52: 1591e1596. Bonagura JD, Miller MW. Doppler echocardiography. II. Color Doppler imaging. Vet Clin North Am Small Anim Pract 1998; 28:1361e1389.

95

34. Thomas WP, Gaber CE, Jacobs GJ, Kaplan PM, Lombard CW, Moise NS, Moses BL. Recommendations for standards in transthoracic two-dimensional echocardiography in the dog and cat. Echocardiography committee of the specialty of cardiology, American college of veterinary internal medicine. J Vet Intern Med 1993;7:247e252. 35. Oyama MA, Sisson DD. Assessment of cardiac chamber size using anatomic M-mode. Vet Radiol Ultrasound 2005;46: 331e336. 36. Rushmer RF, Crystal DK, Wagner C. The functional anatomy of ventricular contraction. Circ Res 1953;1:162e170. 37. Regression Field A. Discovering statistics using IBM SPSS statistics. 4th ed. London: Sage; 2013. p. 293e356. 38. Lindstedt SL, Schaeffer PJ. Use of allometry in predicting anatomical and physiological parameters of mammals. Lab Anim 2002;36:1e19. 39. Schober KE. Doppler echocardiographic assessment of ventricular functionetime to move to the right? J Vet Intern Med 2005;19:785e787. 40. Baumwart RD, Meurs KM, Raman SV. Magnetic resonance imaging of right ventricular morphology and function in boxer dogs with arrhythmogenic right ventricular cardiomyopathy. J Vet Intern Med 2009;23:271e274. 41. Schober KE, Baade H. Doppler echocardiographic prediction of pulmonary hypertension in West Highland white terriers with chronic pulmonary disease. J Vet Intern Med 2006;20: 912e920. 42. Serres F, Chetboul V, Gouni V, Tissier R, Sampedrano CC, Pouchelon JL. Diagnostic value of echo-Doppler and tissue Doppler imaging in dogs with pulmonary arterial hypertension. J Vet Intern Med 2007;21:1280e1289. 43. Hori Y, Ukai Y, Hoshi F, Higuchi S. Volume loading-related changes in tissue Doppler images derived from the tricuspid valve annulus in dogs. Am J Vet Res 2008;69:33e38. 44. Chetboul V, Gouni V, Sampedrano CC, Tissier R, Serres F, Pouchelon JL. Assessment of regional systolic and diastolic myocardial function using tissue Doppler and strain imaging in dogs with dilated cardiomyopathy. J Vet Intern Med 2007; 21:719e730. 45. Baumwart RD, Meurs KM, Bonagura JD. Tei index of myocardial performance applied to the right ventricle in normal dogs. J Vet Intern Med 2005;19:828e832. 46. Chetboul V, Sampedrano CC, Gouni V, Concordet D, Lamour T, Ginesta J, Nicolle AP, Pouchelon JL, Lefebvre HP. Quantitative assessment of regional right ventricular myocardial velocities in awake dogs by Doppler tissue imaging: repeatability, reproducibility, effect of body weight and breed, and comparison with left ventricular myocardial velocities. J Vet Intern Med 2005;19:837e844. 47. Lopez L, Colan SD, Frommelt PC, Ensing GJ, Kendall K, Younoszai AK, Lai WW, Geva T. Recommendations for quantification methods during the performance of a pediatric echocardiogram: a report from the pediatric measurements writing group of the American society of echocardiography pediatric and congenital heart disease council. J Am Soc Echocardiogr 2010;23:465e495. 48. Mawad W, Drolet C, Dahdah N, Dallaire F. A review and critique of the statistical methods used to generate reference values in pediatric echocardiography. J Am Soc Echocardiogr 2013;26:29e37. 49. Cantinotti M, Scalese M, Molinaro S, Murzi B, Passino C. Limitations of current echocardiographic nomograms for left ventricular, valvular, and arterial dimensions in children: a critical review. J Am Soc Echocardiogr 2012;25: 142e152. 50. Cornell CC, Kittleson MD, Della Torre P, Haggstrom J, Lombard CW, Pedersen HD, Vollmar A, Wey A. Allometric

96

L.C. Visser et al.

scaling of M-mode cardiac measurements in normal adult dogs. J Vet Intern Med 2004;18:311e321. 51. Tidholm A, Bodegard-Westling A, Hoglund K, Ljungvall I, Haggstrom J. Comparisons of 2- and 3-dimensional echocardiographic methods for estimation of left atrial size in dogs with and without myxomatous mitral valve disease. J Vet Intern Med 2011;25:1320e1327. 52. Kusunose K, Penn MS, Zhang Y, Cheng Y, Thomas JD, Marwick TH, Popovic ZB. How similar are the mice to men? Between-species comparison of left ventricular mechanics using strain imaging. PLoS One 2012;7:e40061. 53. Bavegems V, Duchateau L, Sys SU, De Rick A. Echocardiographic reference values in whippets. Vet Radiol Ultrasound 2007;48:230e238.

54. Popovic ZB, Sun JP, Yamada H, Drinko J, Mauer K, Greenberg NL, Cheng Y, Moravec CS, Penn MS, Mazgalev TN, Thomas JD. Differences in left ventricular long-axis function from mice to humans follow allometric scaling to ventricular size. J Physiol 2005;568:255e265. 55. Levy PT, Holland MR, Sekarski TJ, Hamvas A, Singh GK. Feasibility and reproducibility of systolic right ventricular strain measurement by speckle-tracking echocardiography in premature infants. J Am Soc Echocardiogr 2013;26:1201e1213. 56. Hall DJ, Cornell CC, Crawford S, Brown DJ. Meta-analysis of normal canine echocardiographic dimensional data using ratio indices. J Vet Cardiol 2008;10:11e23. 57. Geffre A, Friedrichs K, Harr K, Concordet D, Trumel C, Braun JP. Reference values: a review. Vet Clin Pathol 2009;38:288e298.

Available online at www.sciencedirect.com

ScienceDirect