The study population consisted of 20 healthy cats as controls and 17 client-owned cats affected with HCM. All cats presented at Small Animal Teaching Hospital, Faculty of Veterinary Science, Chulalongkorn University, Thailand during August 2018 – June 2019. The study protocol was approved by the Animal Care and Use Committee, Faculty of Veterinary Science, Chulalongkorn University, Thailand (Animal Use Protocol No.1831073). The sample size for each group (at least 17 cats per group) was calculated by the statistical program (G*Power test 3.1) with the estimated standard deviation of the percentage of the LA fractional shortening (LA-FS) from a previous study , an expected 80% power, and a at 0.05, which was enough to find a minimum difference of LA-FS between two groups.
Information on all cats, including breed, sex, age, body weight, body condition score, and clinical findings was recorded. All cats were subject to a complete physical examination, systolic blood pressure measurement and blood collection for complete blood count, blood chemistry and total T4 measurements. Cardiac examination including echocardiography, electrocardiography and thoracic radiography was performed in all cats. No cats had received medications before being enrolled in the study.
Inclusion and exclusion criteria
As inclusion criteria, the control group consisted of healthy cats that had left ventricular wall thickness of < 6 mm and left atrial diameter <16 mm assessed by echocardiography and had no other systemic illness. Cats with left ventricular wall thickness of ≥ 6 mm in at least one region during end diastole  were recruited into the HCM group.
For exclusion criteria, cats affected with other cardiomyopathies including restrictive cardiomyopathy, dilated cardiomyopathy or unclassified cardiomyopathy were excluded. Restrictive cardiomyopathy was identified by left atrial or biatrial enlargement, normal left ventricular wall thickness, normal or decreased systolic function, and restrictive ventricular filling pattern with pulsed wave Doppler echocardiography . Dilated cardiomyopathy was characterized by cardiac chamber dilatation (left ventricular end diastolic dimension > 17 mm) with wall thinning and systolic dysfunction (left ventricular fractional shortening less than 35 % and left ventricular end systolic dimension > 12 mm ) [23-25]. Cats that could not be categorized into the above-mentioned forms of cardiomyopathy were classified as unclassified cardiomyopathy. Cats with renal disease (creatinine >2.0 mg/dL), systemic hypertension (systolic blood pressure >160 mmHg), hyperthyroidism (serum total T4 concentration >4 µg/dl)  and any systemic diseases were excluded from the study.
Two-dimensional and M-mode echocardiography were performed by an investigator (SS). An ultrasound machine (Eko7, Samsung Medison, Seoul, South Korea) with a 4-12 MHz phased array transducer was used. Two-dimensional and M-mode echocardiography was performed on the right parasternal long-axis four-chamber view, to measure the chamber size and wall thickness. The M-mode cursor placed perpendicular to the interventricular septum and left ventricular wall below the tips of the mitral valves at the largest ventricular chamber size. Left ventricular internal dimension at end-diastole and end-systole, interventricular septum thickness at end-diastole and end-systole, left ventricular posterior wall thickness at end-diastole and end-systole were recorded. The ratio of LA to aorta dimension was measured during first diastolic frame of aortic valve closure from a right parasternal short axis view .
Pulsed-wave Doppler and tissue Doppler imaging were used for assessing left ventricular diastolic function. Transmitral flow velocities were measured from the left apical four-chamber view. The gate was placed at the tips of the mitral valve leaflets when they were wide open . Peak velocity of early diastolic transmitral flow, peak velocity of late transmitral flow and the ratio of peak velocity of early diastolic to late diastolic transmitral flow was recorded. Isovolumic (or isovolumetric) relaxation time was measured from the left apical five-chamber view by placing the gate in the left ventricular outflow tract near the anterior mitral valve leaflet to reveal both aortic ejection flow and left ventricular inflow [27, 28]. Pulmonary vein flow velocities were measured in the right parasternal short-axis view . Peak velocity of systolic and diastolic pulmonary vein flow, and flow reversal at atrial contraction were recorded. The ratio of peak velocity of systolic to diastolic pulmonary vein flow were calculated. The myocardial motion along the longitudinal axis of the heart was investigated by placing the gate on the subendocardial portions of the lateral corner of the mitral annulus . Peak velocity of early and late diastolic mitral annular motion and the ratio of peak velocity of early to late diastolic mitral annular motion as determined by pulsed-wave Doppler echocardiography were recorded. The ratio of peak velocity of early diastolic transmitral flow to peak velocity of early diastolic mitral annular motion were calculated .
The LA diameter (LAD) was measured on the right parasternal long-axis four-chamber view parallel with the mitral annulus . The maximal and minimal LAD (LADmax and LADmin) were measured. The LADmax was measured at end-systole (one frame before opening of mitral valve), and the LADmin was measured at peak atrial contraction (one frame before closure of mitral valve) . Changes of the LAD were expressed as the percentage of fractional shortening of the left atrium (LA-FS) by the formula (LADmax-LADmin/LADmax) x100. The left apical four-chamber view was used to measure the maximal LA volume during atrial end-diastole and minimal LA volume during atrial end-systole. The percentage of LA ejection fraction (LA-EF) was calculated by an ultrasound machine automated software . The LA area change (FAC) was measured by tracing the LA endocardial border during LA end diastolic and systolic phases on left apical four- chamber view. Left atrial maximal area (LAAmax) and minimal area (LAAmin) were measured in cm2. Then, FAC was then calculated with the formula FAC = [(LAAmax − LAAmin)/LAAmax] × 100 [5, 7]. The measurements of LA-EF and FAC were performed on the same cardiac cycle as 2D-STE.
Two-dimensional speckle tracking echocardiography (2D-STE)
Two-dimensional speckle tracking echocardiography of the left apical four-chamber view was used to analyze the longitudinal deformation of the left atrium. Two-dimensional echocardiographic images were recorded at 100 frames/s for consecutive three cardiac cycles and three sceneries and stored these in the Digital Imaging and Communications in Medicine format. Offline analysis was performed in images with good quality from each cat. The investigator was blinded to the group where the cats were allocated. The LA wall, including the interatrial septum, the lateral wall and the atrial roof were tracked along during end-diastole. After automatic tracking, manual editing was performed to correct software errors in the region of interest. The ultrasound machine computer software (Strain 2.0 with Bull’s Eye) calculated the LA strain. The mean values of the measurements from three consecutive cardiac cycles were used in all analyses. Six speckle segments were analyzed in each cat. The strain of each segment (as percentages) was plotted on the y-axis versus time (in seconds) on the x-axis over an entire cardiac cycle (Fig. 1). The different colored graphs represent strains from different segments. The white dotted line is the average strain. The peak atrial longitudinal strain (PALS) was manually measured from the peak strain of the average strain value at the end of reservoir phase .
The data from six randomly selected cats in the control group were used for calculating the variability of PALS. For the intra-observer variability, the measurement data from the same operator repeated on two different days (seven days apart) were used. Measurements were performed in the same cardiac cycle from the same cine loop. The inter-observer variability was calculated from measurements of two operators with different levels of experience in echocardiography. The variability was quantified as the coefficient of variation (CV) by the formula, %CV=standard deviation/mean x 100. The degree of repeatability was determined as follows: CV <5%, very low variability; 5-15%, low variability; 16-25% moderate variability; or >25% high variability .
Statistical analyses were performed using a commercially available software (SPSS version 22, Inc, Chicago, IL, USA). Descriptive statistics were used to describe the characteristics of the cats including sex, breed, age, body weight, systolic blood pressure and heart rate. The normality of data was assessed with the Shapiro-Wilks normality test. Comparisons between the two groups (the control and HCM cats) were performed by using the independent student t-test for normally distributed data and the Mann-Whitney U test for non-normally distributed data. The categorical data were compared by using Fisher’s exact test. Comparisons of 3 subgroups, including control and HCM cats with LAD <16 mm, and the HCM cats with LAD >16 mm, were performed by using the one-way Analysis of Variance. The multiple comparisons were performed by the Bonferroni method. An Analysis of Covariance model was used to test the fixed effects of sex, breed, and age, as covariates on conventional and 2D-STE-derived echocardiographic variables. The correlations between PALS and LA-FS, LA-EF, and FAC, assessed by conventional echocardiography, were tested by the Pearson’s correlation coefficient. A value of P <0.05 was considered significant.