• Open Access

Tissue Doppler and Strain Imaging in Dogs with Myxomatous Mitral Valve Disease in Different Stages of Congestive Heart Failure

Authors


  • The examinations were performed at Albano Animal Hospital. The work was presented as an abstract at the 18th ECVIM-CA Congress, September 2008, Ghent, Belgium.

Corresponding author: Anna Tidholm, Albano Animal Hospital, Rinkebyvägen 23, 18236 Danderyd, Sweden; e-mail: anna.tidholm@gmail.com.

Abstract

Background: Tissue Doppler imaging (TDI) including strain and strain rate (SR) assess systolic and diastolic myocardial function.

Hypothesis: TDI, strain, and SR variables of the left ventricle (LV) and the interventricular septum (IVS) differ significantly between dogs with myxomatous mitral valve disease (MMVD) with and without congestive heart failure (CHF).

Animals: Sixty-one dogs with MMVD with and without CHF. Ten healthy control dogs.

Methods: Prospective observational study.

Results: Radial motion: None of the systolic variables were altered and 3 of the diastolic velocities were significantly increased in dogs with CHF compared with dogs without CHF and control dogs. Longitudinal motion: 2 systolic velocities and 3 diastolic velocities were significantly increased in dogs with CHF compared with dogs without CHF and control dogs. Difference in systolic velocity time-to-peak between LV and IVS was significantly increased in dogs with MMVD with and without CHF compared with control dogs. In total, 11 (23%) of 48 TDI and strain variables differed significantly between groups. Left atrial to aortic ratio was positively correlated to early diastolic velocities, percentage increase in left ventricular internal diameter in systole was positively correlated to systolic and diastolic velocities, and mitral E wave to peak early diastolic velocity in the LV basal segment (E/Em) was positively correlated to radial strain and SR.

Conclusions and Clinical Importance: Few TDI and strain variables were changed in dogs with MMVD with and without CHF. Intraventricular dyssynchrony may be an early sign of MMVD or may be an age-related finding.

Abbreviations:
2D

two-dimensional

Ao

aorta

A wave

late diastolic wave

CHF

congestive heart failure

CV

coefficient of variation

DCM

dilated cardiomyopathy

EF

ejection fraction

Em

peak early diastolic velocity in left ventricular basal segment

E wave

early diastolic wave

FS

fractional shortening

HR

heart rate

IVRT

isovolumetric relaxation time

IVS

interventricular septum

LA

left atrium

LV

left ventricle

LVEDP

left ventricular end-diastolic pressure

LVIDd inc %

percentage increase in left ventricular internal diameter in diastole

LVIDs inc %

percentage increase in left ventricular internal diameter in systole

MMVD

myxomatous mitral valve disease

MR

mitral regurgitation

Sa

systolic velocity in mitral annulus

Sm

systolic velocity in left ventricular basal segment

SR

strain rate

TDI

tissue doppler imaging

Myxomatous mitral valve disease (MMVD) is the most common cardiac disease in dogs and many dogs with MMVD eventually develop congestive heart failure (CHF), although at different time periods after disease detection.1

Assessment of left ventricular (LV) systolic and diastolic function is not well established and reasons for development of LV dysfunction and CHF are not clearly understood in patients with mitral regurgitation (MR). In early CHF, slowing of LV relaxation is evident, and diastolic dysfunction is considered to be a primary cause of dyspnea in human patients with CHF.2 As CHF progresses, left atrial pressure is increased with increasing early diastolic filling rate3 and increasing regurgitant volume in MMVD.1 Several variables have been proposed to assess LV systolic function in human patients with severe MR, with ejection fraction (EF) being the most commonly used.4 However, EF remains high during the compensated phase of chronic MR due to decreased afterload and increased preload and sympathetic tone. End-systolic volume or diameter corrected for body weight have been used as more sensitive indices for detecting early myocardial dysfunction in chronic MR.5 Percentage increase in left ventricular internal diameter in systole (LVIDs inc %) was shown to correlate with worse outcome in dogs with MMVD and CHF.6 However, LV internal dimensions and EF reflect hemodynamic consequences of both MR overload and myocardial function, whereas tissue Doppler-derived indices are considered relatively independent of loading conditions.2 Tissue Doppler imaging (TDI) has been extensively evaluated in healthy dogs and in dogs with different heart diseases.7–13 This relatively new technique allows quantitative assessment of segmental and global myocardial radial and longitudinal motion in systole and diastole, as well as accurate timing of events in relation to the QRS complexes on ECG recordings. TDI variables have been reported in dogs with MMVD and concomitant pulmonary hypertension.12 Early myocardial diastolic dysfunction in human patients with LV volume overload may be detected using TDI velocities in the longitudinally aligned subendocardial fibers.14 Noninvasive assessment of LV diastolic pressure by peak transmitral E wave divided by peak early diastolic velocity in the LV basal segment (E/Em) was reported to be a strong predictor of adverse outcome in human patients with severe MR, as was systolic intraventricular dyssynchrony.2,15 Strain and strain rate (SR) are TDI-derived variables that measure tissue deformation and rate of tissue deformation, respectively. SR is defined as the difference of tissue velocities per myocardial segment length, and strain is the integral of SR.13,16 Radial strain and SR are positive in systole when the analyzed segment undergoes lengthening, and SR shows 2 negative waves during diastole, when shortening occurs. Longitudinal strain and SR are negative during systole, and SR shows 2 positive waves during diastole. Studies of strain and SR have been performed in normal dogs17 and in dogs with dilated cardiomyopathy (DCM).8

The purposes of this study were as follows: (1) to investigate whether or not tissue Doppler and strain imaging variables differ between MMVD dogs with and without CHF and (2) to investigate possible correlations between different variables and estimates of left atrial volume overload as assessed by left atrium-to-aortic ratio (LA/Ao), myocardial systolic function assessed by percent increase in LVIDs inc %, and LV filling pressure assessed by E/Em.

Materials and Methods

Animals and Procedures

Dogs were included in the study based on the following criteria: (1) presence of echocardiographic evidence of MMVD, (2) TDI velocity, strain, and SR tracings of sufficient quality for adequate analysis. Ten healthy dogs of similar weight were also examined using the same equipment and the same protocol. These dogs were considered healthy on the basis of complete physical examination, ECG, conventional echocardiography, and Doppler examinations. Thoracic radiography was performed in all dogs with clinical signs of CHF such as cough, dyspnea, or exercise intolerance. All radiographs were examined by one veterinary specialist in cardiology (AT) and the cardiac silhouette, pulmonary parenchyma, and vessels were assessed.

Classification of CHF

The classification of CHF used in this study is based on the CHIEF system which is a modified system based on that proposed by the American College of Cardiology and the American Heart Association.18–20 This classification system appears to be superior to previously used systems, because it provides a continuum of stages with and without therapy ranging from A to D, with class A denoting risk for heart disease and class D denoting end-stage CHF. Class B includes animals with structural heart disease but no clinical signs of CHF. In class BI cardiomegaly is mild or absent radiographically, whereas cardiomegaly is present in class BII. Class C includes animals with clinical and radiographic signs of left- or right-sided CHF, with animals not yet receiving therapy classified as CIII, and animals receiving therapy for CHF with signs of CHF decreased or absent as class CII. Class DIV includes animals in refractory end-stage CHF. In this study, echocardiographic measurements of the cardiac chambers were used for class B animals to differentiate between class BI and BII. In the statistical analysis, dogs without CHF included class BI and BII dogs, and dogs with CHF included class CII, CIII, and DIV dogs.

Conventional Echocardiography and Doppler Examination

Conventional 2-dimensional (2D) and M-mode echocardiographic and Doppler examinations were performed by 1 experienced veterinary specialist in cardiology (AT) with an ultrasound unita equipped with 3.0–8.5 MHz phased-array transducers in all dogs. Dogs were unsedated and gently restrained in left and right lateral recumbency during the examination. Measurements were made using the 2D-guided M-mode with concomitant ECG-registration for the ventricles according to the American Society of Echocardiography.21 Percent increase in LVID and LVIDs was calculated as follows: % increase = [100 × (observed dimension−expected normal dimension)/expected normal dimension].6 Expected normal dimensions were calculated according to the following method: expected normal LVIDd = 1.53 × (BW)0.294; expected normal LVIDs = 0.95 × (BW)0.315.22 Measurements of Ao and LA were made on the 2D parasternal short-axis view obtained at the level of the aortic valve.23 Mitral regurgitant flow velocities were determined from a left apical 4-chamber view in all dogs and mitral inflow velocities where early (E) and late (A) diastolic flow could be distinguished.

TDI and Strain Imaging

2D color TDI examinations were performed by the same experienced veterinary specialist in cardiology (AT) with the same ultrasound unit as used for standard echocardiography with concomitant ECG registration. Real-time color Doppler images were superimposed on 2D images with a frame rate ≥ 150 frames/s, and the Doppler velocity range was set as low as possible to avoid aliasing artifacts (Figs 1–4). The region of interest was positioned between the papillary muscles for the short-axis view with a width of 0.5 cm and a length extending from the endocardium to the epicardium. In the longitudinal view, the region of interest was placed within the interventricular septum (IVS) and the LV, respectively, with a width of 0.5 cm and a length extending from the apical region to the mitral valve annulus, taking care to precisely follow the myocardium. Each myocardial wall was recorded separately to enable the smallest sector width possible. An automated tracking system was used to ensure that the sampling region would stay within the myocardial wall during the recording. In the short-axis as well as the long-axis view, the myocardium was divided into 4 segments, and velocity and strain tracings were recorded simultaneously. Velocity tracings of the radial endocardial and epicardial segments, and the longitudinal basal and apical segments were used in the analyses. Strain and SR were only analyzed for the mid-wall segment in both axes, because recordings of the other 3 segments generally were considered of inadequate quality. Strain and SR may have a high signal-to-noise ratio, especially with increased length of the sampling segment. After completion of all echocardiographic examinations, off-line measurements were made using a software program.b The operator was blinded to the patient's status during the off-line analyses. Depending on the heart rate (HR), 3–5 consecutive cardiac cycles were available for analysis. Measurements were averaged for each variable and mean values were used in the statistical analysis. Ten to 15% of dogs considered for the study were excluded based on insufficient quality of TDI images, either at the time of image acquisition or at the time of analysis.

Figure 1.

 Radial tissue doppler velocity curves of 4 myocardial segments in a control dog. S, systolic velocity; E, E wave velocity; A, A wave velocity.

Figure 2.

 Longitudinal interventricular tissue doppler velocity curves of 4 myocardial segments in a dog with myxomatous mitral valve disease. S, systolic velocity; E, E wave velocity; A, A wave velocity.

Figure 3.

 Radial midwall strain in a dog with myxomatous mitral valve disease. S, systolic strain.

Figure 4.

 Radial midwall strain rate in a dog with myxomatous mitral valve disease. S, systolic wave; E, E wave velocity; A, A wave velocity.

Radial LV free wall variables resulting from radial motion were measured using the right parasternal short-axis view between the papillary muscles. Peak velocities for the subendocardial and subepicardial segment and velocity gradients, defined as the difference between subendocardial and subepicardial velocities, were determined in systole, and in early and late diastole. Peak systolic strain and SR was measured for the mid-wall segment. Time-to-peak, defined as the time period from the beginning of the R wave on the ECG to the peak of the waveform, was measured for systolic velocities, strain, and SR.

Longitudinal LV free wall and IVS variables resulting from longitudinal motion were measured from the left apical 4-chamber view. Each wall was recorded separately to increase the frame rate and enhance the quality of the recording. Basal and apical peak velocities and velocity gradient, defined as the difference between the basal and apical velocities, were determined in systole, and in early and late diastole. Peak early diastolic velocity in the LV basal segment (Em) was used for calculation of transmitral E/Em as a filling pressure index.15 Peak systolic strain and SR were determined for the mid-wall segment of the LV wall and IVS. Time-to-peak was measured for systolic velocities, strain and SR, and the time difference between the LV wall and the IVS systolic velocities was calculated as an index of ventricular synchrony.

Assessment of Repeatability

Within-day variability was assessed using 6 dogs, including 3 dogs without cardiac disease and 3 dogs with MMVD (2 class BII and 1 class CII). Each dog was examined 6 times on a given day, and 25 TDI and strain variables were recorded (Table 1). Each TDI and strain imaging variable was measured on 3–5 consecutive cardiac cycles on the same frame, and the resulting mean values and standard deviations were used to determine the coefficient of variation (CV). TDI velocities generally had lower CV values (<15% for all measured velocities and time intervals except 1) in comparison with strain and SR for both groups. The lowest within-day variability was found in the systolic velocity of the basal longitudinal segments and in the radial endocardial segment. The highest CV values were obtained for the difference in time-to-peak for the LV and IVS. CV values <15% were obtained for 72% of all variables for dogs without cardiac disease compared with 61% of all variables for dogs with MMVD.

Table 1.   Within-day variability of tissue doppler derived velocities, strain and strain rate (SR) for both radial and longitudinal motions of the left ventricle (LV) and the interventricular septum (IVS) in 6 dogs (3 dogs without cardiac disease and 3 dogs with MMVD (2 class BII and 1 class CII).
VariableSDCV (%) and
Range
  1. CV, coefficient of variation; MMVD, myxomatous mitral valve disease; SD, standard deviation.

Radial endocardial systolic velocity0.296.2 (2–8)
Radial endocardial E wave velocity0.7312 (6–18)
Radial endocardial A wave velocity0.4914.3 (12–19)
Radial systolic time-to-peak5.75 (2–9)
Radial strain8.120 (12–31)
Radial strain time-to-peak3212.5 (9–24)
Radial SR0.9519 (12–39)
Radial SR time-to-peak13.514 (6–24)
Longitudinal IVS basal systolic velocity0.256.8 (2–11)
Longitudinal IVS basal E wave velocity0.459.8 (7–19)
Longitudinal IVS basal A wave velocity0.4011.5 (4–22)
Longitudinal IVS systolic time-to-peak8.59 (8–12)
Longitudinal IVS strain5.422 (11–34)
Longitudinal IVS strain time-to-peak25.311.6 (6–17)
Longitudinal IVS SR1.124 (18–37)
Longitudinal IVS SR time-to-peak10.511.6 (6–16)
Longitudinal LV basal systolic velocity0.327.7 (2–16)
Longitudinal LV basal E wave velocity0.4210 (4–18)
Longitudinal LV basal A wave velocity0.2812.6 (6–26)
Longitudinal LV systolic time-to-peak8.98 (4–13)
Longitudinal LV strain0.6515.7 (11–22)
Longitudinal LV strain time-to-peak27.99.7 (3–20)
Longitudinal LV SR0.6514.8 (6–23)
Longitudinal LV SR time-to-peak412 (5–22)
Difference in LV to IVS systolic time-to-peak5.243.8 (23–73)

Statistical Analysis

A computer programc was used for all statistical analyses. A Kruskal-Wallis test was used for testing equality of medians among the 3 groups of dogs. For variables in which the medians were significantly different (P < .05), a pair-wise comparison between the groups also was performed using Mann-Whitney U-test with Bonferroni's adjustment, in which a P value <.017 was considered significant. The associations between LA/Ao and LVIDs inc % and TDI variables were investigated using Spearman's rank correlation. Values are reported as medians and interquartile ranges.

Results

Seventy-one dogs of 21 different breeds were prospectively included in the study: Cavalier King Charles Spaniel (26), Dachshund (11), Bichon Frisé (4), mixed breed (4), Chinese Crested Powder Puff (3), Tibetian Terrier (3), Border Collie (2), English Springer Spaniel (2), Miniature Schnauzer (2), Norfolk Terrier (2), Nova Scotia Duck Tolling Retriever (2), and 1 of 10 other small- to medium-sized breeds. According to the CHIEF classification, 32 dogs were classified without CHF (30 with class BI and 2 with class BII) and 29 dogs were classified with CHF (22 with class CII, 6 with class CIII, and 1 with DIV). Dogs with CHF were treated with furosemide (29), benazepril (22), enalapril (2), pimobendan (6), digoxin (5), and spironolactone (4). Ten dogs were healthy controls. Forty-two dogs (59%) were males and 29 dogs (41%) were females. Weight ranged from 3 to 29 kg (median, 10 kg). There were no statistically significant differences between dogs with and without CHF and healthy controls concerning sex or body weight. Age ranged from 13 to 187 months and was significantly lower (P < .001) in healthy control dogs compared with dogs with MMVD with and without CHF (median 10 years for both groups with MMVD and 5.6 years for control dogs). Median HR was significantly higher (P < .0001) in dogs with CHF compared with dogs without CHF and control dogs (140, 115, and 100 beats/min, respectively).

Conventional Echocardiography and Doppler Examination

In dogs with CHF, percentage increase in left ventricular internal diameter in diastole (LVIDd inc %), LVIDs inc %, LA/Ao, mitral E wave peak velocity, and E/A were significantly higher compared with dogs without CHF and normal control dogs. Fractional shortening (FS) and mitral E/Em did not differ significantly among groups (Table 2).

Table 2.   Median values and IQR for clinical and conventional echocardiography and Doppler variables in dogs with myxomatous mitral valve disease with (n = 29) and without (n =32) CHF and healthy controls (n =10).
VariableDogs with CHFDogs without CHFControl DogsOverall P Value
  1. Values with different superscript letters indicate statistically significant differences between groups.

  2. CHF, congestive heart failure; FS, fractional shortening; IQR, interquartile range; LA/Ao, left atrial to aortic diameter ratio; LVIDd inc %, percentage increase in left ventricular internal diameter in diastole; LVIDs inc %, percentage increase in left ventricular internal diameter in systole; MR, mitral regurgitation.

Age (years)10 (9–12)a10 (8–11)a5.6 (3–8)b<.0001
Body weight (kg)9.3 (8–11)a10 (8.6–12.6)a14 (8–20)a.14
Heart rate (beats/min)140 (120–165)a115 (100–125)b100 (92–127)b<.0001
LVIDd inc %44.0 (34–62)a10.6 (2.4–28)b−1 (−10–7)b<.0001
LVIDs inc %22.6 (−0.8–44)a5.5 (−11–19)b−6.7 (−19–8)b.0015
LA/Ao1.72 (1.5–2.1)a1.07 (1–1.2)b1.02 (0.9–1.1)b<.0001
FS (%)45.7 (40–49)a39.9 (35–45)a39.8 (35–43)a.06
Mitral E wave (m/s)1.2 (0.9–1.5)a0.8 (0.6–0.9)b0.6 (0.5–1)b.0013
Mitral A wave (m/s)0.8 (0.7–0.9)a0.7 (0.6–0.8)a0.7 (0.5–0.7)a.06
Mitral E/A1.4 (1.1–2.1)a1.2 (1–1.4)b1.0 (0.9–1.5)b.03
Mitral E/Em21 (19–31)a18 (15–22)a18 (12–32)a.29
Velocity of MR jet (m/s)4.7 (4–5)a4.7 (4–5.3)a.47

TDI and Strain Imaging

There were no statistically significant differences in radial systolic velocity, strain or SR of the LV free wall among groups. In dogs with CHF, the radial epicardial early diastolic velocity and the endocardial E/A were significantly increased compared with dogs without CHF and normal controls. The radial late diastolic gradient was significantly decreased in dogs with CHF compared with dogs without CHF and normal control dogs (Table 3). Longitudinal IVS early diastolic apical velocity was increased in dogs with MMVD with and without CHF compared with control dogs, and basal E/A was increased in dogs with CHF compared with dogs without CHF and control dogs. Longitudinal LV systolic basal and apical velocities were significantly higher in dogs with CHF compared with dogs without CHF and normal controls, as was early diastolic apical velocity. The difference in time-to-peak between LV and IVS systolic waves, despite similar HRs between recordings, was significantly greater in dogs with and without CHF compared with control dogs, indicating dyssynchrony. Time-to-peak LV and IVS systolic strain was significantly shorter in dogs with CHF compared with dogs without CHF and control dogs (Tables 4 and 5).

Table 3.   Median values and IQR for radial tissue Doppler and strain imaging variables in dogs with myxomatous mitral valve disease with (n = 29) and without (n = 32) CHF and healthy controls (n = 10).
VariableDogs with CHFDogs without CHFControl DogsOverall P Value
  1. Values with different superscript letters indicate statistically significant differences between groups.

  2. CHF, congestive heart failure; IQR, interquartile range.

Endocardial systolic velocity (cm/s)6.0 (4.6–6.7)a5.6 (4.2–6.1)a5.4 (4.7–7.6)a.33
Epicardial systolic velocity (cm/s)3.5 (3–4.9)a3.2 (2.4–3.9)a3.1 (2.5–4.4)a.25
Velocity gradient (cm/s)2.2 (1.4–2.8)a2.0 (1.4–2.5)a2.3 (2.9–5.3)a.15
Systolic time-to-peak (ms)115 (98–126)a110 (96–122)a101 (85–122)a.56
Endocardial E wave velocity (cm/s)4.6 (3.9–5.7)a4.0 (3–4.9)a3.9 (2.9–5.3)a.14
Epicardial E wave velocity (cm/s)3.4 (2.8–4.4)a1.6 (1.4–2.7)b2.4 (1.9–3)b.0006
E wave gradient (cm/s)1.4 (0.5–2.7)a1.8 (1.2–2.6)a1.5 (0.9–2.7)a.45
Endocardial A wave velocity (cm/s)2.5 (1.9–3.8)a3.0 (2.3–3.7)a3.8 (3.5–5.3)a.08
Epicardial A wave velocity (cm/s)1.6 (0.9–2.6)a1.5 (1.1–2.3)a2.0 (1.4–2.6)a.44
A wave gradient1.1 (0.8–1.4)a1.4 (0.7–1.8)a2.1 (1.6–2.3)b.03
Endocardial E/A1.8 (1.3–2.4)a1.4 (0.9–1.8)a1.1 (1–1.5)b.01
Epicardial E/A1.8 (1.2–4.7)a1.1 (0.9–1.7)a1.2 (0.9–2.1)a.11
Mid–wall strain (%)53 (45–63)a52 (40–58)a39 (33–48)a.20
Mid-wall strain time-to-peak (ms)215 (199–239)a228 (198–257)a223 (201–276)a.77
Mid-wall strain rate (s−1)6.0 (5–10.8)a5.7 (4–6.9)a4.2 (3.6–5.6)a.30
Mid-wall strain rate time-to-peak (ms)100 (87–110)a96 (78–108)a126 (89–140)a.14
Table 4.   Median values and IQR for longitudinal interventricular tissue Doppler and strain imaging variables in dogs with myxomatous mitral valve disease with (n = 29) and without (n = 32) CHF and healthy controls (n = 10).
VariableDogs with CHFDogs without CHFControl DogsOverall P Value
  1. Values with different superscript letters indicate statistically significant differences between groups.

  2. CHF, congestive heart failure; IQR, interquartile range.

Basal systolic velocity (cm/s)6.1 (5–7)a4.9 (4.3–6.1)a5.0 (4–5.9)a.14
Apical systolic velocity (cm/s)3.0 (2.3–3.5)a2.3 (1.9–2.8)a1.9 (1.4–2.7)a.03
Systolic velocity gradient (cm/s)3.2 (2.2–3.8)a2.6 (2–3.5)a3.1 (2–3.9)a.64
Systolic time-to-peak (ms)82 (70–89)a92 (77–108)a94 (84–114)a.049
Basal E wave velocity (cm/s)4.6 (3.8–6)a4.0 (3.5–4.7)a4.0 (2.4–4.7)a.42
Apical E wave velocity (cm/s)2.5 (2.2–4.6)a2.4 (1.5–3)a1.8 (0.7–2.7)b.03
E wave velocity gradient (cm/s)2.0 (1.4–2.5)a1.6 (1.1–2.3)a1.8 (0.7–2.7)a.75
Basal A wave velocity (cm/s)4.4 (3.1–5.1)a4.2 (3.5–4.9)a3.5 (3.3–4.2)a.31
Apical A wave velocity (cm/s)2.0 (1.3–2.9)a1.8 (1.5–2.5)a1.8 (0.5–2.2)a.68
A wave velocity gradient (cm/s)2.2 (1.5–3.1)a2.0 (1.5–3.4)a1.7 (1.1–3.1)a.55
Basal E/A1.2 (1.1–1.8)a1(0.8–1.2)b1.1 (0.9–1.5)b.03
Apical E/A1.4 (1.2–2.1)a1.1 (0.8–1.8)a1.6 (0.8–2.6)a.16
Mid-wall strain %−26 (−23–−35)a−25.0 (−21–−32)a−21.7 (−19–−32)a.35
Mid-wall strain time-to-peak (ms)236 (202–250)a256 (240–276)b278 (164–288)b.008
Mid-wall strain rate (s−1)−2.8 (−2–−3.9)a−2.6 (−2.1–−3.1)a−2.1 (−1.6–−2.4)a.24
Mid-wall strain rate time-to-peak (ms)83 (73–102)a87 (79–104)a101 (82–111)a.22
Table 5.   Median values and IQR for longitudinal left ventricular free wall tissue Doppler and strain imaging variables in dogs with myxomatous mitral valve disease with (n = 29) and without (n = 32) CHF and healthy controls (n = 10).
VariableDogs with CHFDogs without CHFControl DogsOverall P Value
  1. Values with different superscript letters indicate statistically significant differences between groups.

  2. CHF, congestive heart failure; IQR, interquartile range.

Basal systolic velocity (cm/s)6.0 (5–6.8)a4.7 (3.7–5.5)b4.4 (3.6–5.8)b.003
Apical systolic velocity (cm/s)2.7 (2.3–3.4)a2.1 (1.5–2.7)b1.6 (0.8–3.5)b.01
Systolic velocity gradient (cm/s) (Sm)3.1 (2–4.3)a2.5 (1.9–3.6)a2.9 (2.1–3.7)a.41
Systolic time-to-peak (ms)101 (88–111)a100 (88–116)a95 (83–118)a.93
Difference in LV to IVS systolic time-to-peak (ms)17 (9–33)a11 (2–26)a2 (−5–6.5)b.04
Basal E wave velocity (cm/s) = Em4.9 (3.5–7)a4.2 (3.2–4.9)a4.5 (3.4–5.8)a.25
Apical E wave velocity (cm/s)3.0 (2.6–4.1)a2.2 (1–2.7)b1.4 (1–2.5)b.006
E wave velocity gradient (cm/s)1.8 (1–2.9)a2.4 (1–2.8)a3.1 (2.2–3.8)a.07
Basal A wave velocity (cm/s)3.0 (1.9–4.1)a3.1 (2.3–4)a3.6 (2.1–5.5)a.44
Apical A wave velocity (cm/s)1.3 (0.9–1.7)a0.9 (0.5–1.3)a0.7 (0.5–1.1)a.16
A wave velocity gradient (cm/s)1.7 (1.1–2.3)a2.2 (1.6–2.9)a2.4 (1.2–4.7)a.25
Basal E/A1.5 (1.1–2.4)a1.3 (0.9–1.8)a1.1 (0.9–1.4)a.31
Apical E/A2.2 (1.7–3.6)a2.0 (1.4–3)a1.9 (1–3.4)a.72
Mid-wall strain (%)−17(−15–−22)a−21 (–15–−26)a−22 (−15–−29)a.16
Mid-wall strain time-to-peak (ms)231 (200–250)a257 (240–272)b278 (255–290)b.001
Mid-wall strain rate (s−1)−1.9 (−1.5–−2.9)a−2.0 (−1.2–−3.2)a−2.3 (−1.9–−2.8)a.72
Mid-wall strain rate time-to-peak89 (81–99)a100 (85–114)a91 (79–117)a.33

Bivariate Analyses

Left-sided volume overload, assessed by LA/Ao, was positively correlated to HR, LVIDd inc %, LVIDs inc %, mitral E wave peak velocity, E/A, E/Em, radial epicardial E wave velocity and gradient, longitudinal IVS basal E/A, E wave velocity and gradient, and LV basal systolic velocity. LA/Ao was negatively correlated to radial E wave gradient, time-to-peak for the longitudinal IVS and LV strain (Table 6).

Table 6.   Bivariate analysis of significant correlations of different variables with LA/Ao in dogs with myxomatous mitral valve disease with and without CHF.
VariableSpearman's ρP Value
  1. Ao, aorta; Em, peak early diastolic velocity in LV basal segment; IVS, interventricular septum; LA, left atrium; LV, left ventricle; LVIDd inc %, percentage increase in left ventricular internal diameter in diastole; LVIDs inc %, percentage increase in left ventricular internal diameter in systole.

Heart rate0.41.0008
LVIDd inc %0.72<.0001
LVIDs inc %0.52<.0001
Mitral E wave peak0.61<.0001
Mitral E/A0.44.0006
E/Em0.39.022
Radial epicardial E wave velocity0.55<.0001
Radial E wave velocity gradient−0.32.031
Longitudinal IVS basal E wave velocity0.32.041
Longitudinal IVS E wave velocity gradient0.39.012
Longitudinal IVS basal E/A0.52.0004
Longitudinal IVS strain time-to-peak−0.37.0002
Longitudinal LV basal systolic velocity0.35.010
Longitudinal LV strain time-to-peak−0.39.012

Myocardial systolic function assessed by LVIDs inc % was positively correlated to LA diameter, LA/Ao, mitral E wave peak velocity, E/A, radial systolic and early diastolic epicardial velocity, longitudinal IVS basal and apical E/A, and LV systolic basal velocity and gradient. LV end-systolic percentage increase was negatively correlated to FS (Table 7).

Table 7.   Bivariate analysis of significant correlations of different variables with percentage increase in left ventricular internal diameter in systole (LVIDs inc %) in dogs with myxomatous mitral valve disease with and without CHF.
VariableSpearman's ρP Value
  1. FS, fractional shortening; IVS, interventricular septum; LA/Ao, left atrial to aortic diameter ratio; LVIDd inc %, percentage increase in left ventricular internal diameter in diastole; LVIDs inc %, percentage increase in left ventricular internal diameter in systole.

FS−0.52<.0001
LA0.57<.0001
LA/Ao0.52<.0001
LVIDd inc%0.81<.0001
Mitral E wave peak0.53<.0001
Mitral E/A0.41.003
Radial epicardial systolic velocity0.31.018
Radial epicardial E wave velocity0.44.003
Longitudinal LV basal systolic velocity0.34.015
Longitudinal LV systolic velocity gradient0.31.026
Longitudinal IVS basal E/A0.60<.0001
Longitudinal IVS apical E/A0.39.019

LV filling pressure assessed by E/Em was positively correlated to HR, LA/Ao, radial mid-wall strain and SR, and longitudinal IVS E wave gradient. Transmitral E to Em ratio was negatively correlated to Ao diameter, longitudinal IVS systolic velocity time-to-peak, and LV apical E wave velocity (Table 8).

Table 8.   Bivariate analysis of significant correlations of different variables with mitral E/Em in dogs with myxomatous mitral valve disease with and without CHF.
VariableSpearman's ρP Value
  1. CHF, congestive heart failure; Em, peak early diastolic velocity in LV basal segment; IVS, interventricular septum; LA/Ao, left atrial to aortic diameter ratio.

Heart rate0.34.047
Ao−0.37.029
LA/Ao0.39.022
Radial strain0.49.010
Radian train rate0.55.009
Longitudinal IVS E wave velocity gradient0.40.033
Longitudinal IVS systolic velocity time-to-peak−0.61.0003
Longitudinal LV apical E wave veloctiy−0.42.014

Discussion

The present study shows that few (23%) systolic and diastolic tissue Doppler and strain variables were altered in dogs with MMVD and CHF compared with dogs with MMVD without CHF and control dogs. For those variables that were different between the groups, the systolic and diastolic tissue velocities were increased and time intervals were shorter in dogs with CHF, which is in agreement with classical conventional echocardiographic findings of severe MMVD (ie, shortened systolic time intervals and hyperkinesia).24 In a recent study of 110 dogs of various breeds with MMVD, systolic and diastolic TDI and strain variables were increased in small breed dogs with moderate disease, and decreased in small and large breed dogs with severe disease.25

HR was significantly increased in dogs with CHF compared with dogs without CHF and control dogs in the present study. This finding is expected in all studies comparing dogs with and without CHF due to the basic pathophysiology of CHF.26 Therefore, it is diffucult to evaluate whether or not differences found in the present study of different TDI and strain variables between dogs with and without CHF are due to increased HR or to other factors.

Systolic myocardial dysfunction as indicated by significantly greater LVIDs inc % was present in dogs with MMVD and CHF compared with dogs with MMVD without CHF and to control dogs in the present study. FS did not differ significantly among groups, which is in agreement with previous studies showing that this variable is a comparatively less sensitive indicator of systolic function.27,28 LV internal dimensions as well as EF and FS reflect hemodynamic consequences of both volume overload and myocardial function, whereas TDI and strain imaging have been considered to be less load-dependent techniques. Early myocardial dysfunction might be detected in the subendocardial fibers in patients with LV volume overload.14 As these fibers are aligned longitudinally, alterations of LV basal (Sm), or mitral annulus (Sa), systolic velocities may be expected. Although several investigators reported lower velocities with myocardial systolic dysfunction in human patients with MR,14,29,30 1 study reported significantly increased systolic annular velocities in human patients with primary MR compared with patients without MR,31 which is in agreement with the findings of the present study. Teshima et al32 reported no difference between dogs with MMVD with and without CHF in Sa. Systolic velocities of the LV free wall and the IVS were significantly increased in dogs treated with dobutamine, and significantly decreased in dogs treated with esmolol compared with baseline in a recent study of normal dogs.13 Increased longitudinal systolic velocities compared with baseline were reported for dogs with MMVD without CHF treated with pimobendan.33 In the present study, LV basal systolic velocities were significantly increased in dogs with CHF compared with dogs without CHF and control dogs. Because only 6 dogs (10%) in our study were treated with pimobendan and 5 additional dogs (8%) were treated with digoxin, it is unlikely that the findings in this study were influenced by treatment with inotropic drugs. A more likely reason would be that TDI estimates of myocardial performance were influenced by an increased sympathetic drive and the basic pathophysiology of MR (ie, volume overload in conjunction with a comparatively low afterload because of the ejection of blood into the atrium).

Diastolic myocardial function is commonly assessed by mitral valve inflow velocities using conventional Doppler techniques. LV diastolic filling pressures are invariably increased in CHF.2,34 In our study, mitral E and A wave velocities as well as E/A were significantly increased in dogs with CHF compared with dogs without CHF and control dogs, which is in agreement with previous studies in dogs.10,35 In contrast to transmitral velocities which are affected by preload, afterload, HR as well as by LA and LV compliance, TDI variables appears to be less dependent on loading conditions.36 Among the TDI diastolic variables, the longitudinal LV basal (Em), or mitral annulus (Ea) velocity is particularly useful as it represents myocardial relaxation and may be used as an index of early ventricular diastolic function. Mitral E/Em has been shown to correlate well with pulmonary capillary wedge pressure (PCWP) in several studies of human patients and can be used to estimate LV filling pressures.2,15 E/Em has also been shown to be a strong prognosticator in human studies, especially E/Em ≥ 15 and Em ≤ 3 cm/s.37 In a study of naturally occurring MMVD in dogs, E/Em was significantly increased in dogs with CHF compared with dogs without CHF.31 E/Em was significantly increased from baseline and showed a strong correlation with mean left atrial pressure in an experimental study of dogs with acute MR.38 LV end-diastolic pressure (LVEDP) was directly measured in a recent study of dogs with pacing-induced CHF, where E/IVRT ratio was useful, but E/Em was not, in predicting decrease in LVEDP induced by furosemide administration.10 These findings are in agreement with our study, where no difference in E/Em was found among groups, regardless of supposedly different filling pressures. As preload dependency of Em was shown in an experimental study of dogs, making the use of E/Em as an index of filling pressures inaccurate,39 this finding is not surprising. Interestingly, E/Em was found to be a reliable estimate of LV filling pressures only in human patients with significant secondary, but not primary, MR.31 In our study, E/Em ratio generally was higher compared with that reported in other studies in dogs, which may partly be explained by the fact that myocardial velocities obtained from the on-line pulsed wave TDI curve are higher than those in our study which were reconstructed from 2D color-coded TDI images off-line, thus decreasing the ratio. Breed-related differences also may influence the results.15,40,41

In the present study, E/Em was significantly correlated to LA/Ao, which is in agreement with a previous study of dogs with MMVD,32 as well as with HR and radial strain and SR. In the present study, Em did not differ significantly among groups, which is contrary to findings in a study of dogs with MMVD in which dogs with CHF had significantly lower Em velocities compared with dogs without CHF.32 However, in our study several other diastolic variables, such as radial epicardial E wave, radial endocardial E/A, longitudinal IVS apical E wave and basal E/A, and LV apical E wave velocities, were significantly increased in dogs with CHF compared with dogs without CHF and control dogs, which is in agreement with an experimental study of dogs with acute MR.38 A paradoxically faster mitral annular velocity during early diastole was found in human patients having LV myocardial dysfunction with moderate to severe MR and considerably high LV filling pressures.42 The reason for this apparently improved diastolic function in dogs and humans with MR and CHF is unclear. The age of the dogs in the control group was significantly lower compared with dogs with MMVD with or without CHF. In a study of dogs with MMVD, Em was not significantly correlated with age,32 whereas Em has been reported to be inversely correlated with age in human patients.43

In the present study, IVS systolic time-to-peak was significantly shorter in dogs with CHF compared with dogs without CHF and control dogs. This finding could be explained by increased sympathetic drive in conjunction with decreased resistance for LV emptying because of MR. Intraventricular dyssynchrony, measured as difference in time-to-peak between LV and IVS, was significantly increased in dogs with MMVD with and without CHF compared with control dogs despite of a rather high CV. Myocardial dyssynchrony thus may be considered an early sign of MMVD in dogs or may be an age-related finding, as control dogs were of significantly lower mean age compared with MMVD dogs. Longitudinal systolic and diastolic velocity curves were reported to be highly synchronized in a study of healthy humans, in whom age and HR predominantly affected diastolic, but not systolic, variables.44 Myocardial disease is reported to cause delay in time of onset of Sm,45 and Sm time-to-peak was significantly prolonged in hypertrophic, hypertensive, and DCM in human patients.46,47 Patients with intraventricular dyssynchrony were, independent of QRS width and EF, at significantly higher risk of cardiac events,48,49 and LV systolic and diastolic mechanical dyssynchrony is reported to be common in patients with CHF.50 Systolic velocity time-to-peak was longer in the lateral wall compared with that of IVS in dogs with DCM.9

Strain is a TDI-derived modality assessing myocardial deformation. As the ventricle contracts, the myofibers shorten (negative strain) in the longitudinal direction and lengthen (positive strain) in the radial direction. SR measures the rate of deformation and seems to be correlated to the rate of change in pressure (dP/dT), which reflects contractility. In contrast to TDI velocity data, which reflect movement of 1 tissue site relative to the transducer, strain and SR reflect the movement of a tissue site relative to another within the sample volume.51 Strain imaging has been shown to be a repeatable and reproducible method for assessing systolic myocardial function in healthy dogs, relatively independent of age and HR.28 In a study of dogs with DCM, systolic strain was significantly lower compared with control dogs.8 In our study, there were no significant differences among groups regarding strain or SR in either radial or longitudinal motion, which may partly be explained by the relatively high CV. LV and IVS mid-wall strain time-to-peak was significantly shorter in dogs with CHF compared with dogs without CHF and control dogs. Radial strain and SR were positively correlated with E/Em, and LV strain time-to-peak was negatively correlated to LA/Ao.

Limitations of Study

Technical aspects of recording and analyzing TDI-derived variables must be considered, because the reliability of the measurements may be questioned especially in remodelled hearts, and where excessive motion of the heart occurs. We found that the CV was acceptable for most velocity variables, but was comparatively higher for strain and SR variables, because strain and SR are more angle dependent and sensitive to noise. Coefficients of variation also were higher in dogs with CHF compared with dogs without CHF, as would be expected due to increased movement of the heart and the thorax. However, analyzing CV only for dogs without heart disease, as is commonly reported, may not be valid for studies including dogs with heart disease.

Lack of an age-matched control group is a limitation of this study. However, it is difficult to find an age-matched control group without some degree of MR, and if dogs with MR are used as control dogs in the study, it may be difficult to define what degree of MR will be considered trivial and not clinically relevant. Only 1 variable (ie, intraventricular dyssynchrony) measured as difference in time-to-peak between LV and IVS, was significantly increased in dogs with MMVD with and without CHF compared with control dogs. Additional studies are needed to investigate whether or not this finding is primarily age related or related to presence of MMVD or both.

The present study indicates that few TDI variables were changed in dogs with CHF caused by MMVD compared with compensated MMVD dogs. The variables that differed significantly among the groups were in most cases highly co-variate with other variables obtained from conventional echocardiography, suggesting that they contribute little additional information. These findings suggest 1 or both of the following: (1) myocardial systolic and diastolic functions are comparatively well preserved in dogs with CHF caused by MMVD, (2) loading conditions and sympathetic tone affect TDI and strain imaging variables to a greater extent than previously expected, suggesting that TDI imaging is associated with the same problems as conventional echocardiography in evaluation of systolic and diastolic dysfunction.

Footnotes

aHD XE 11, Philips Ultrasound, Bothell, WA

bQLAB advanced quantification, Philips Ultrasound

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

Ancillary