Pulsed-wave Doppler tissue imaging velocities in normal geriatric cats and geriatric cats with primary or systemic diseases linked to specific cardiomyopathies in humans, and the influence of age and heart rate upon these velocities

Abstract

Pulsed-wave Doppler tissue imaging (pw-DTI) techniques allow the non-invasive assessment of myocardial dynamics. pw-DTI has demonstrated regional and global diastolic impairment in various forms of human and feline cardiomyopathy. We hypothesise that in geriatric cats with systemic diseases that have been linked to specific cardiomyopathies in human beings, the myocardial velocity profile will be altered when compared to either normal or hypertrophic cardiomyopathy (HCM) cats; and that both age and heart rate have a significant affect upon pw-DTI velocities. The aims of this study were to determine whether the feline M-mode or myocardial velocity profile is altered in geriatric cats with disease states that have been linked to specific cardiomyopathies in humans when compared to normal geriatric cats or geriatric cats with HCM and to determine whether age or heart rate has a significant effect upon pw-DTI velocities within these groups of cats. Sixty-six cats aged 8 years or above were included in the study, and were divided as follows: Unaffected (n=8), basilar septal bulge (BSB) (17), HCM (14), hyperthyroid (HiT₄) (12) and chronic renal failure (CRF) (15). Systolic blood pressure was normal in all the cats. pw-DTI systolic (S′), early (E′) and late diastolic (A′) velocities were assessed from standardised sites within the myocardium, and the relationships between these and disease group, age and heart rate were then assessed. In cats with HCM, the E′ velocity was decreased at various sites. Conversely, the HiT₄ cats demonstrated increased S′ velocities. The only site at which the age of the cat was significantly related to myocardial velocities was the S′ velocity from the apical mid-septum. There were also significant positive relationships between heart rate and the magnitude of myocardial S′, E′ and A′ velocities of radial motion and S′ and A′ velocities of longitudinal motion. pw-DTI detected diastolic dysfunction in untreated cats with HCM and increased systolic function in HiT₄ cats. The age of the cat was of little significance, whereas heart rate significantly influenced myocardial velocity profiles.

The World Health Organisation defines cardiomyopathies as diseases of the myocardium associated with cardiac dysfunction. ¹ In addition, it states that hypertrophic cardiomyopathy (HCM) is characterised by left and/or right ventricular hypertrophy, which is usually asymmetric and involves the interventricular septum and, that the left ventricular volume is typically normal or reduced. ¹ However, this classification system also distinguishes between ‘primary’ cardiac hypertrophy (HCM) and cardiac dysfunction occurring secondary to systemic disease, for which the term ‘specific cardiomyopathy’ is ascribed. The specific cardiomyopathies are further divided according to the underlying aetiology. ¹

This human classification system has been broadly applied to the feline cardiomyopathies, with all of the principle ‘primary’ cardiomyopathies that have been seen in humans having been demonstrated in cats. ² Furthermore, some of the specific cardiomyopathies that have been reported in human beings, are also documented in cats, namely thyrotoxic cardiomyopathy and hypertensive cardiomyopathy. ^3–6 To date, in the cat, the existence of a specific ‘uraemic cardiomyopathy’ has not been reported. However, there have been a number of reports of specific cardiac abnormalities occurring in cats with chronic renal failure (CRF). ^7–11 For the purpose of this study, the term ‘specific cardiomyopathy’ will be used to collectively refer to cats with cardiac changes secondary to hyperthyroidism, hypertension, CRF or diabetes mellitus (DM).

Cardiac function can be accurately assessed by invasive measurements, however, these techniques are not widely available, nor are they applicable to clinical cases. Doppler tissue imaging (DTI) techniques have been developed to provide an alternative tool for the non-invasive quantification of regional and global myocardial function. ^12–14 In humans, these DTI derived variables correlate well with invasive measures of systolic and diastolic function. ¹³ However, there appears to be both an age- and heart rate- related effect, with diastolic function deteriorating with age ^13,15 and the ratio of the late to early diastolic velocity increasing with heart rate. ¹⁶

Pulsed-wave DTI (pw-DTI) measures the instantaneous peak myocardial velocity as it moves through a pre-selected sampling gate, and in people with myocardial disease it is a relatively load-insensitive technique. ^17,18 Early diastolic velocity is decreased in people with HCM. ¹⁹ In addition, in humans, genetically altered rabbits and Maine Coon cats with a genotype discriminating for HCM, pw-DTI has detected diastolic dysfunction prior to the onset of hypertrophy. ^20–22 DTI indices can also be used to assess feline myocardial motion velocity. ^6,23–26 Peak systolic and diastolic velocities recorded from the mitral annulus are decreased in cats with HCM or hypertension, when compared to normal. ^23,25,27 However, to date, the velocity of myocardial motion has not been reported in any of the other potential feline specific cardiomyopathies.

Pulsed-wave DTI variables have been shown to demonstrate a significant correlation with invasive measures of diastolic function in healthy anaesthetised cats. ²⁴ Heart rate acts as an independent variable, affecting pw-DTI derived A′ velocities and with this the ratio of E′:A′. In contrast, a recent study which assessed colour DTI (c-DTI) velocities in a group of healthy cats, found a positive correlation between S′ and E′ (or summated E′+A′) velocities and heart rate, at various sites within the myocardium. ⁶ To date, studies have focused only on the effects of heart rate on DTI velocities recorded from healthy cats. There have, as yet, been no studies assessing the effects of heart rate on DTI velocities in cats with either HCM or disease states linked to specific cardiomyopathies.

A weak, but statistically significant, relationship between pw-DTI velocities and age has been previously reported in a study of 25 normal cats. ²⁵ This relationship was lost in cats with HCM (n=23). ²⁵ However, there were three cats in the normal group less than 1 year old, and it is unclear whether or not the relationship is continuous throughout life, or is a developmental change with little clinical relevance in adult cats. The diseases associated with feline specific cardiomyopathy, tend to be diseases of geriatric cats. If comparisons are to be made between normal cats and cats with specific cardiomyopathies the relationship between age and myocardial dynamics warrants further investigation.

The aims of this study were, therefore, to determine whether the feline myocardial velocity profile is altered in geriatric cats with various systemic diseases which have been associated with specific cardiomyopathies in humans, and to compare these profiles to the myocardial velocity profiles recorded from normal geriatric cats or geriatric cats with HCM. In addition, we aimed to determine whether either age or heart rate has a significant effect upon these velocities.

Materials and Methods

Case selection

This study comprises part of a large cross-sectional study was performed over a 3-year period on 134 cats geriatric cats (aged 8 years or older). ⁵ All of the cats were privately owned animals, selected from cases seen at either the first opinion or referral clinics of the University of Edinburgh, Hospital for Small Animals, and lived within the Edinburgh area. All owners gave informed written consent. A full history was taken, with particular attention being paid to any history of cardiac disease, duration of illness, and current or previous medications.

Blood pressure measurement

Once admitted to the hospital, the cats were allowed approximately 1 h to acclimatise, and were then taken to a quiet room where systolic blood pressure was recorded using an indirect Doppler flow-metre (Parks Electronic Doppler Flow-Meter; model B-811, Perimed, Bury St Edmunds, UK). The technique used was consistent with guidelines set out by the hypertension consensus panel. ²⁸

Once blood pressure had been recorded, a full physical examination was performed. Particular attention was paid to hydration status, the presence of thyroid nodules, renal size and shape, the presence of hepatomegaly and the pulse quality.

Systemic disease identification

Routine serum biochemistry and haematology profiles, including total thyroxine (T₄), were assessed for all cats (University of Edinburgh Veterinary Pathology Unit. Haematology analysis was performed using a Pentra 60 haematology analyser (ABX Diagnostics, Montpellier, France). Biochemistry was performed on an opeRA autoanalyser (Bayer Diagnostics, Tarrytow, NY, USA). Electrolytes were measured using an AVL 9180 Electrolyte Analyser (AVL Scientific Corporation, Roswell, USA)). (Cardiovit AT-60, Schiller AG, Switzerland). Cats with a blood glucose concentration of 145.6 mg/dl (8 mmol/l), or greater, had a serum fructosamine level measured.

Any azotaemic cats that had not previously been diagnosed with renal insufficiency, had a repeat blood sample taken approximately 2 weeks later for re-assessment of serum urea and creatinine concentrations.

Cats were placed in right lateral recumbency and a six-lead electrocardiogram (ECG; GE Medical Systems-Horten, Norway) was recorded. The amplitude, duration and timing of complexes were measured, and compared with reference values. ²⁹ A note was made of any rhythm or conduction abnormalities. Any cat found to be in atrial fibrillation or with a significant conduction abnormality was excluded from the study.

Echocardiographic examination

Echocardiographic examination was performed using a Vingmed Vivid FiVe ultrasound machine with built-in DTI capacity equipped with a 5 MHz flat phased array (FPA) probe (Verbatim Optical, Japan). All images were stored digitally on optical discs (GE Medical Systems for Macintosh) and analysed retrospectively. Whilst scanning, an ECG trace (lead II) was simultaneously recorded. Sector width and depth were adjusted to maximise cardiac image. Gain settings were optimised for each case, and focus adjusted to the area of interest. Pulsed Doppler tracings were made using high pulsed-repetition frequency (HPRF) Doppler modality. For each variable the mean of five measurements from adequate, consecutive, waveforms on a single trace was taken.

All images were acquired with the cats in lateral recumbency, scanning through the dependent chest wall. All cats were scanned unsedated by a single operator (KES).

Two-dimensional and M-mode evaluation

From the right parasternal long-axis view the interventricular septal diameter was measured at end-diastole using the leading edge-to-leading edge technique (Fig 1) from just below the level of the mitral valve, and perpendicular to the long-axis of the heart.

Fig 1.

The arrow represents the point at which the thickness of the interventricular septum was assessed. The boxes represent the location for cursor placement in the pw-DTI analysis. (a) Right parasternal long-axis view (b) right parasternal short-axis view (c) left apical four-chamber view.

The M-mode recording taken from the right parasternal short-axis, at chordae level, was used to measure the left ventricular chamber dimensions again using the leading edge-to-leading edge technique. ³⁰ End-diastole was identified by the beginning of the Q wave on the ECG; systole was measured at the point of maximal septal excursion. ³

A two-dimensional, short-axis right-sided parasternal image of the aorta and left atrium was used to measure the aortic and left atrial diameters at end-diastole using an inner-edge to inner-edge technique. The aortic diameter was measured from the mid-part of the right coronary cusp to the commissures of the left and non-coronary cusps. The left atrial diameter was measured in the same frame, by extending the line used to measure the aorta; care being taken to avoid the pulmonary vein. ³¹

Doppler tissue imaging

The techniques used to obtain the pw-DTI profiles have been described previously. ²⁶ Briefly, velocity profiles were recorded from the mid-ventricular level of the free wall imaged from the right parasternal long- and short-axis views (hereafter referred to as the long- and short-axis free wall); from the septal and lateral aspects of the mitral annulus imaged from the left apical four-chamber view (hereafter referred to as the apical septal and apical lateral mitral annulus) and at the mid-ventricular level of the interventricular septum (the apical mid-septum) and free wall imaged from the left apical four-chamber view (apical mid-free wall) (Fig 1). For all the pw-DTI analyses a sampling gate was set at 3.9 mm.

Echocardiographic post-processing

Images were analysed using the Echopac software (GE medical Systems for Macintosh). All measurements were performed by a single investigator (KES). Measurement of the pw-DTI was performed a minimum of 2 weeks after the scan was performed. Diastolic events were analysed at end-diastole, which was identified by the start of the QRS complex. ³ Atrial systole was identified by its association with the P wave. ³² Systolic events were measured at end-systole, which was defined by the end of the T wave. ³ Events associated with any premature or ectopic beats were not measured; the complexes immediately prior to, and after, any ectopic complexes were also ignored. For each measurement, the associated R–R interval was measured and the mean R–R interval was calculated, from which the mean heart rate was then calculated. The horizontal axis of the tracings was magnified to increase the accuracy of the timing and duration measurements.

Myocardial excursion patterns

For all the pw-DTI recordings the principal velocity during systole (S′), and the two principal velocities during diastole (E′ and A′) were measured. Occasionally the E′ and A′ waves were fused; in these cases the E′+A′ velocity was measured and recorded as such.

Cat groups

Cats were classified as ‘unaffected’ cats if they were clinically healthy, normal on physical examination. In addition, routine blood sampling revealed no significant abnormalities, systolic blood pressure was between 120 and 160 mmHg, auscultation, a six-lead ECG and standard two-dimensional and M-mode echocardiographic examinations revealed no abnormalities.

The remaining cats were grouped according to their systolic blood pressure, haematology and serum biochemistry results, and echocardiography findings. Cats were classified as hyperthyroid (HiT₄) if their serum T₄ was ≥3.72 μg/dl (48 nmol/l), laboratory reference range: 1.02–3.72 μg/dl (13–48 nmol/l). ³³ CRF cats had serum creatinine exceeding 2.0 mg/dl (177 μmol/l), laboratory reference range: 0.45–2.0 mg/dl (40–177 μmol/l) ³⁴ on two consecutive occasions, at least 14 days apart. DM cats had blood glucose exceeding 145.6 mg/dl (8 mmol/l), laboratory reference range: 59.5–90 mg/dl (3.3–5.0 mmol/l) and serum fructosamine above 350 μmol/l, laboratory reference range: 100–350 μmol/l. ³⁵ Systemic hypertension was diagnosed if systolic blood pressure exceeded 200 mmHg, or was between 170 and 200 mmHg with concomitant ocular lesions, or was over 170 mmHg on two or more subsequent visits. ²⁸ Any cat with a history of HCM prior to the onset of any of these systemic diseases was excluded from the study. Cats were diagnosed with HCM if they did not fall into any of the above groups, the haematocrit was within normal limits (packed cell volume 0.24–0.45 l/l), and there was M-mode evidence of concentric hypertrophy ² (either the left ventricular free wall or interventricular septum measuring 6 mm or greater at end-diastole, on standard M-mode). ³⁶ Cats were excluded from the study if they demonstrated chamber dilation or evidence of other forms of acquired or congenital cardiac disease. During the study period we identified a group of cats demonstrating a localised area of hypertrophy affecting only the basilar septum, with no other cardiac abnormalities. The group was classified as having a lone basilar septal bulge (BSB) (Fig 2). There were no changes in serum biochemistry results, packed cell volume was within the reference range, systolic blood pressure was <170 mmHg, and M-mode measurements and left atrial diameter were within normal limits. ³⁶ Cats with more than one of the diseases outlined above (eg, hypertension and CRF), or any systemic disease that may influence myocardial dynamics, were excluded from the analysis. All cats receiving medication were excluded from the analysis.

Fig 2.

Right parasternal long-axis view demonstrating a BSB.

Statistical analysis

The statistical analysis was carried out using R (V2.4.1, The R development core team). Differences in the proportion of cats that were female between groups were estimated by χ² analyses. The age, systolic blood pressure and T₄ between groups were compared using Kruskal–Wallis as this data could not be adequately normalised via transformation (eg, log). Dwass–Steel–Critchlow–Fligner's multiple comparison post-hoc procedure for non-parametric data was used to establish which groups were different from each other. ³⁷

One-way analyses of variance were used to compare differences between groups in: heart rate at presentation; respiratory rate; blood pressure; serum creatinine concentration; serum total T₄ at initial presentation; and heart rates during echocardiographic acquisition of each of the six pw-DTI variables. Two creatinine values were excluded from the analysis as they were >2.6 times as large as any other values. All parameters required Box–Cox transformation prior to analysis in order to normalise the residuals. One-way analyses of variance were also used to examine the relationship between cat groups and the standard echocardiographic M-mode parameters and the two-dimensional assessment of left atrial diameter. In all cases, post-hoc multiple comparisons of the mean values of the above parameters in the groups using Tukey contrasts were undertaken if any overall statistically significant differences between groups were found.

For each myocardial velocity that was assessed a single analysis of covariance (ANCOVA) was carried out to look at the influence of group, age and heart rate. The ANCOVAs were carried out with the interaction between group and age and group and heart rate included. If either of the interactions was not statistically significant an ANCOVA excluding them was used. All the velocities required Box–Cox transformation prior to analysis in order to normalise the residuals. If there were any statistically significant differences between groups in particular velocities post-hoc multiple comparisons of the mean values using Tukey contrasts were undertaken. A P-value of <0.05 was taken as demonstrating statistical significance, and the degrees of freedom associated with particular test statistics are quoted as subscripts.

Results

Overall, 134 cats were assessed for inclusion. Of these cats, 35 were excluded from the analysis as they were suffering from more than one disease, and a further 32 because they were receiving medication at the time of scanning. All of the cats with DM (n=10) or acromegaly (n=4) were receiving medications, as were all but one of the hypertensive cats (5/6). Therefore, all of these disease groups were excluded from the analysis.

The final study population comprised 66 cats, split in the following five groups: eight were judged to be free of disease (unaffected); 17 BSB, 14 asymptomatic HCM, 12 hyperthyroid (HiT₄), and 15 CRF. Fifty-seven of the 66 cats (86%) were either domestic shorthair or longhair (DSH/DLH).

Population characteristics

The median age was 11.5 years (range 8–18). Thirty were neutered females and 36 neutered males, with no gender bias in any disease group (all χ²₁<2.1, P>0.149). However, there was a gender bias in the CRF DSH/DLH cats (75% male, χ²₁=5.3, P=0.022). HiT₄ cats were significantly older than the unaffected and BSB cats (P<0.027, Fig 3a). The mean weight of the 64 cats (which had weight recorded) was 4.75 kg (range 2.15–6.60). The HiT₄ cats were significantly lighter than HCM cats (F_4,59=2.68, P=0.040).

Fig 3.

Boxplots comparing (a) age at presentation (years); (b) serum creatinine concentration (μmol/l); (c) heart rate at the apical mid-septum (seconds, s); (d) heart rate at the long-axis free wall (s); (e) heart rate at the short-axis free wall (s); (f) heart rate at the apical mid-lateral wall (s); (g) heart rate at the apical lateral mitral annulus (s); and (h) heart rate at the apical septal annulus (s) for the different cat groups. In all cases the boxes represent the interquartile range, the white horizontal bar the median value for that group, and the whiskers the range of data for that group. Creatinine and all heart rate data are plotted on Box–Cox transformed axes. Hatched areas represent the laboratories reference range. The horizontal whiskers indicate statistically significant groups as assessed by post-hoc multiple comparisons of the mean values in the groups using Tukey contrasts.

Mean heart rate at initial presentation was 177 bpm (range 140–220), with no statistically significant difference between groups (F_4,61=1.81, P=0.138). In the 49 cats for which the respiratory rate was recorded at initial presentation, the mean respiratory rate was 36 bpm (range 20–80). There was no statistically significant difference between groups (F_4,44=2.58, P=0.050). The median systolic blood pressure was 148 mmHg (range 100.8–160 mmHg); with one of the hyperthyroid cats being mildly hypotensive, again there were no statistically significant differences between groups (P=0.107).

In contrast, the T₄ concentrations did differ between groups (P<0.001), with not surprisingly, HiT₄ cats having significantly higher T₄ than any of the other groups of cats. In addition, there were statistically significant differences in serum creatinine concentrations between groups (F_4,59=25.0, P<0.001, Fig 3b), with CRF cats having higher and HiT₄ having lower creatinine concentrations than the other groups.

Heart rate

When the heart rate (assessed by measurement of the R–R interval during the acquisition of the pw-DTI variables) was estimated it was noted that the heart rate tended to be highest in the HiT₄ group and lowest in the unaffected group (Fig 3c–h). However, the only statistically significant difference found was for the heart rate assessed at the apical mid-septum (F_4,61=2.69, P=0.039, Fig 3c; other five heart rates all F<2.32, P>0.066).

Echocardiographic examination

Analysis of the standard echocardiographic examination demonstrated that the thickness of both the left ventricular posterior wall, in systole and diastole, and the interventricular septum, in both systole and diastole differed significantly between groups (F>4.86, P<0.002), with HCM cats having thicker walls than unaffected and BSB cats in all four measurements (Fig 4a–d). In addition, HCM cats had statistically thicker walls than CRF cats for left ventricular posterior wall, in systole and diastole, and the interventricular septum in systole, with the end maintained in the interventricular septum in diastole (Fig 4d). Finally BSB cats had thinner left ventricular posterior walls in diastole compared to HiT₄ cats (Fig 4b), with this trend observed in the other three measurements. None of the groups varied significantly in any of the other routine echocardiographic parameters assessed (all F<1.9, P>0.122).

Fig 4.

Boxplots comparing (a) left ventricular thickness in systole (cm); (b) left ventricular thickness in diastole (cm); (c) interventricular septum in systole (cm); and (d) interventricular septum in diastole (cm). In all cases the boxes are as for Fig 3. The horizontal whiskers indicate statistically significant groups as assessed by post-hoc multiple comparisons of the mean values in the groups using Tukey contrasts.

Myocardial velocities

Overall 396 Doppler tissue imaging traces were assessed. Only, 14 demonstrated summation of the E and A waves, which were excluded from the analysis. Figure 5 presents the velocities observed at the six locations at the three stages subdivided into cat group. The ANCOVA analyses revealed that the only statistically significant differences between groups in their S′ velocities were in the apical septal annulus (F_4,58=2.96, P=0.027, Fig 5p) with HiT₄ cats having faster velocities than CRF. This trend was also observed for the other five S′ velocities (Fig 5 a,d,g,j,m).

Fig 5.

Boxplots comparing systolic (a,d,g,j,m,p), early-dystolic (b,e,h,k,n,q) and last-dystolic (c,f,i,l,o,r) velocities (m/s) as measured at the long-axis free wall (a–c); short-axis free wall (d–f); apical mid-lateral wall (g–i); apical lateral mitral annulus (j–l); apical mid-septum (m–o); and apical septal annulus (p–r) different cat groups. In all cases the boxes are as for Fig 3. All myocardial velocities are plotted on Box–Cox transformed axes. The stripe shaded box represents previously reported values for normal cats (mean±1 standard deviation), ²⁵ with no previously reported normal values for the short-axis wall velocities. The horizontal whiskers indicate statistically significant groups as assessed by post-hoc multiple comparisons of the mean values in the groups using Tukey contrasts.

Three of the six E′ velocities (short-axis free wall, apical lateral mitral annulus and apical mid-septum) displayed statistically significant differences between groups (all F>2.66, P<0.042, Fig 5 e,k,n), with HCM cats having slower velocities than the unaffected cats in all three. This trend was also observed in the other three velocities (all P<0.110, Fig 5b,h,q). In addition, HCM cats had significantly slower E′ velocities at the apical lateral mitral annulus compared to HiT₄ cats, and at the apical mid-septum compared to BSB cats, with both trends observed in at the other locations. Finally, the apical mid-septum E′ velocities of CRF cats were slower than that of unaffected cats.

Only one of the six sites had statistically significant variation in the A′ velocities – the long-axis free wall (F_4,5=4.82, P=0.002, Fig 5c), with HiT₄ cats having faster velocities than BSB and HCM cats, however, this trend was not repeated for the other locations (Fig 5f,i,l,o,r).

Myocardial velocities and age

Four of the six S′ velocities (long-axis free wall, apical lateral mitral annulus, apical mid-septum, apical septal annulus) had negative relationships with age. However, only one (apical mid-septum) emerged from the ANCOVA analysis as statistically significant (F_1,57=4.15, P=0.046, other three F<3.26, P>0.076), and the fit of the relationship was poor (R²=11.5). The other two locations (short-axis free wall and apical mid-lateral wall) had non-significant positive relationships with age (F<3.1, P>0.083). All six E′ velocities had negative relationships with age, however, none was statistically significant (all F<3.4, P>0.070). No real pattern was observed for the A′ velocities and age with three non-significant positive (short-axis free wall, apical mid-septum and apical septal annulus (all F<3.1, P>0.083)), and three non-significant negative relationships (long-axis free wall, apical mid-lateral wall and apical lateral mitral annulus (all F<0.36, P>0.554)). There were no statistically significant interactions between age and disease group for any of the velocities at any of the sites.

Myocardial velocities and heart rate

All six S′ velocities had a positive relationship with heart rate; the relationship statistically significant for long-axis free wall, apical mid-lateral wall and apical mid-septum (all P<0.015), but the tightnesses of fit were poor (R²<0.117). With the exception of apical lateral mitral annulus, all E′ velocities also had a positive relationship with heart rate, with the relationship statistically significant for long-axis and short-axis free wall (both P<0.046), but again the tightnesses of fit were poor (R²<0.124). Finally all six A′ velocities had a positive relationship with heart rate, with the relationship statistically significant for all measurements apart from apical septal annulus (P other measurements <0.043), but again the tightness of the fits were poor (R²<0.062). The only statistically significant interaction between heart rate and cat group to emerge from the ANCOVA analyses was in terms of the A′ velocity measured at the apical septal annulus (P=0.027), with HCM cats having a statistically different positive velocity and heart rate relationship, compared to the negative relationships for unaffected and HiT₄ cats.

Discussion

This is the first study to assess the myocardial dynamics in a range of feline disease states which have been linked to specific cardiomyopathies in human beings, and then compare these findings to normal cats and cats with HCM.

Population characteristics

The cats included in this study demonstrated few deviations in their baseline characteristics. One of the characteristics that did demonstrate significant variation was the age of the cats, with the HiT₄ cats being significantly older than the unaffected cats and cats with BSB. Analysis of the effect of age upon velocity demonstrated no differences between the different disease groups, therefore, the slight difference in the age of the groups at baseline, does not appear to have significantly affected the velocities reported here.

As was determined by the inclusion criteria, the serum T₄ and creatinine concentrations were higher in the cats with hyperthyroidism and CRF, respectively. However, the serum creatinine concentration was lower in cats with hyperthyroidism compared to the other groups of cats. It has previously been reported that hyperthyroid cats are prone to weight loss and an increased catabolic state which causes muscle wastage and so decreases the total body creatinine concentration. ³⁸ The hyperthyroid cats did have a tendency towards a decreased body weight. In addition, a decreased serum creatinine concentration has previously been described in cats with thyrotoxicosis. ^38,39 This is thought to result from an increased glomerular filtration rate. ^38,39 Either, or both, of these mechanisms may be responsible for the decreased serum creatinine concentrations in the hyperthyroid cats in the current study.

Heart rate

The cats with HiT₄ tended to have higher heart rates than the unaffected cats. This was only significant when the apical mid-septum was assessed. The elevated heart rate (during the echocardiographic examination) suggests that the cats within the HiT₄ group demonstrated increased sympathetic tone during this period. ⁴⁰

Myocardial velocities

The HiT₄ cats demonstrated an increase in the S′ velocities at the majority of sites. These velocities were significantly higher within this group when measured at the apical septal annulus. Previously, hyperthyroid cats have demonstrated an increased fractional shortening, which may be caused by an increased inotropic state. ⁴ The pw-DTI analysis provided evidence for an increase in systolic function at multiple sites within the myocardium. There are several reasons why hyperthyroid cats might demonstrate an increase in inotropic state; these include up-regulation of the sympathetic nervous system, as has been demonstrated by increased numbers of β₁-adrenoreceptors in hyperthyroidism, ⁴¹ or an increased metabolic demand. ⁴² In humans, pw-DTI S′ velocities are positively correlated with β-adrenoceptor density, and negatively correlated with the percentage of interstitial fibrosis. ⁴³ Therefore, one could hypothesise that the increased S′ velocities in the hyperthyroid cats may be related to an increased β-receptor density in the hyperthyroid state.

The E′ velocities tended to be lower in the cats with HCM than in the unaffected cats at all the sites assessed. This reached statistical significance when assessed at the apical mid-septum and at the lateral mitral annulus and the short-axis free wall. Previously, a decreased peak diastolic pw-DTI velocity has been demonstrated in cats with HCM compared to normal cats, ^23,35,27 and that this dysfunction is evident in the longitudinal fibres to a greater extent than it is in the radial fibres. ²⁵ This study demonstrates that diastolic dysfunction is present within the longitudinal and radial fibres of cats with HCM.

This study is the first to demonstrate diastolic dysfunction in cats with CRF. This group of cats tended to demonstrate decreased E′ velocities compared to the unaffected cats. This decrease reached significance at the apical mid-septum. This group has previously demonstrated that acquisition of E′ from this point is more repeatable than acquisition from other sites. ²⁶ There has previously been some suggestion that cats with CRF may demonstrate cardiac abnormalities. ^7–11 None of the cats included in the current study were anaemic or suffering from hypertension. Therefore, decreased E′ velocities within the longitudinal motion recorded at the mid-septal level supports a hypothesis diastolic dysfunction in aged cats with CRF.

In man, it has been demonstrated that there is diastolic dysfunction in many patients with CRF prior to the commencement of dialysis. ⁴⁴ In addition, histological studies have identified myocardial interstitial fibrosis in people suffering from uraemia. This fibrosis was independent of hypertension, DM, anaemia, heart weight and the presence or absence of dialysis. ⁴⁵ It is recognised that many cats with ventricular hypertrophy have myocardial interstitial fibrosis, ⁴⁶ but to date, no studies have looked for an association with renal insufficiency.

Relationship between myocardial velocities and age

In the current study age had a minimal effect on the pw-DTI velocities in either a normal or diseased geriatric cat population. Although there were relatively few unaffected cats within the sample population, there was no evidence that pw-DTI velocities varied with age in this group, aged 8–14 years. Furthermore, the reported velocities are within one standard deviation of the pw-DTI velocities reported previously. ²⁵ In that study the normal group was comprised 25 cats with a mean age of 6±3.5 years. Previously this group has reported that in normal cats the majority of the pw-DTI measurements demonstrate a co-efficient of variability of <20%. ²⁶ The differences between these two studies are well within the inherent variability of this technique. In light of these results, it appears that no correction factor is required when normal geriatric cats are assessed.

When all of the cats were analysed (both unaffected and diseased), it was found that only the S′ velocity recorded from the apical mid-septum was significantly associated with the age of the cat. However, the tightness of the relationship was poor. None of the other 17 velocities examined demonstrated a significant relationship with age.

In human beings, as individual's age, left ventricular diastolic function is impaired. ⁴⁷ It is believed that this occurs in part, due to impaired myocardial relaxation and in part due to increased myocardial stiffness (myocardial fibrosis and collagen deposition increasing with age). ⁴⁷ In humans, it has been demonstrated that there is a significant negative correlation between E′ and age and a positive correlation between A′ and age. ^15,47 However, in children, the relationship between pw-DTI velocities and age appears to differ from that within the adult population, an increase in early diastolic velocities in the first few years of life, followed by a positive correlation with age being reported. ^48,49

In cats, we have previously reported a significant negative association between pw-DTI E′ velocities and age in a group of 25 normal cats aged 10 months to 14 years (mean 6±3.5 years). ²⁵ In contrast, the current study identified no significant association between the pw-DTI E′ velocity and age at the majority of sites assessed in either the unaffected geriatric cats, or the diseased cats. However, in the previous study we reported that there was no relationship between age and pw-DTI E′ velocities in cats with HCM. Whilst the previous study was performed on cats of any age, it included, predominantly, young to middle age cats (three were aged 1 year, or less, and only three were over 9 years of age). The current study investigated the effect of age on pw-DTI velocities in cats over 8 (8–18) years of age. Therefore, there was minimal overlap between the age distributions studied. The cumulative results of these studies suggest that age does not have a significant influence on the myocardial velocities recorded from cats with cardiac hypertrophy.

Relationship between myocardial velocities and heart rate

A significant relationship between many pw-DTI variables and the heart rate was observed. The strongest association was observed in the measurements obtained from the right parasternal long- and short-axis views (assessing the radial myocardial velocities). All of the measurements taken from the long-axis, and both the E′ and A′ measured from the short-axis, demonstrated significant variations with heart rate, although the R² values were relatively low, with heart rate accounting for a maximum of 12.4% of the variability. In comparison, when assessing the longitudinal function only the S′ and A′ velocities were consistently affected by heart rate, with the E′ velocities demonstrating no significant variation with heart rate. Therefore, it appears that the variation is heart rate is more pronounced in the radial fibres, affecting both the systolic and diastolic velocities. These findings imply that as the heart rate changes the heterogeneity of contraction alters. ²⁵

The effects of heart rate upon DTI parameters have been investigated previously. An invasive study has demonstrated that the A′ velocity of longitudinal motion was related to heart rate in the cat. ²⁴ Another study using c-DTI techniques in healthy cats reported a significant relationship between heart rate and both S′ and E′ along the longitudinal axis. ⁶ However, in that study 67% of the diastolic velocities were summated and, therefore, A′ was not assessed. Instead, the summated E′ and A′ were measured as E′. This may have affected the results of the E′ analysis. Furthermore, the effect of heart rate on pw-DTI variables in cats with a primary cardiomyopathy has been assessed. ²³ In that study, heart rate significantly influenced diastolic measurements. A similar finding is reported in the current study.

The current study was limited primarily by the small sample size; this was especially true for the unaffected group of cats, which comprised only eight individuals. Unfortunately, geriatric cats without systemic or cardiac disease were not readily identified within the population seen at our hospital. A concerted effort was made to recruit additional cases to this group. However, the majority of cats recruited as presumed ‘unaffected’ did in fact demonstrate a dynamic murmur associated with BSB.

Despite these limitations, the reported findings suggest that in geriatric cats, irrespective of disease status, pw-DTI velocities are generally not related to age. Therefore, it is not necessary to correct for age when using pw-DTI to study myocardial function in this age group. In addition, this study also demonstrates that heart rate significantly influences multiple pw-DTI velocities, primarily those assessing radial motion. However, the effect of heart rate upon longitudinal function appears minimal. This supports the concept of heterogeneous contraction, suggesting a greater variation in the radial motion with heart rate. In light of these findings, the influence of heart rate upon Doppler variables should always be considered in the interpretation of pw-DTI variables.

In the aged cat population studied, the basilar interventricular septum was frequently greater than 6 mm. As previously reported E′ velocities recorded from this group of cats were lower than those recorded from the unaffected cats, ⁵⁰ however, this did not reach statistical significance. It has been hypothesised that the localised area of hypertrophy seen in these cats may be either a form of senile remodelling, or a localised from of HCM. Further studies are warranted to investigate whether or not the myocardial dynamics in cats with BSB differ significantly from either normal cats or cats with HCM.

In the group of cats with HCM, we demonstrated a significant decrease in the E′ velocity recorded by pw-DTI, at the short-axis free wall, the apical mid-interventricular septum, and at the apical lateral aspect of the mitral annulus. Diastolic dysfunction has previously been described in this group. ^23,25 As any cat receiving medication was excluded from this study, these findings demonstrate that pw-DTI techniques are able to identify diastolic dysfunction in a group of cats with asymptomatic HCM.