Two-dimensional speckle tracking echocardiography (STE) is an advanced imaging technique useful to assess myocardial function by strain and strain rate analysis in human and veterinary medicine [1
]. This novel echocardiographic modality assesses myocardial deformation using the standard 2-D images to follow the speckles (natural acoustic tissue reflections) frame-by-frame during the cardiac cycle [12
]. Using STE, clinicians can evaluate myocardial deformation independent from translational cardiac movements and beam angle [1
]. Several studies have focused on assessing the components of myocardial deformation of the left ventricle (LV), using the left apical four-chamber (LAP4Ch) view for longitudinal strain and strain rate and right parasternal short-axis views for radial and circumferential deformation in dogs and cats [4
Recently, investigators have described the feasibility of assessing longitudinal strain and strain rate using STE from the right parasternal four-chamber (RP4Ch) view, rather than the LAP4Ch view in dogs [18
]. This approach provides a high quality four chamber image for STE analysis, avoiding the limits of LAP4Ch view (i.e., tracking errors due to breathing or side lobe artefacts, especially for LV apex and free wall) [18
]. However, the authors of that study reported that a suboptimal visualization of the apical segments during the cardiac cycle from the RP4Ch view can induce tracking errors and data obtained from this view and LAP4Ch view were not interchangeable [18
]. As the authors stated [18
], RP4Ch is often more easily obtained than the LAP4Ch view in dogs; this is true also for feline patients.
To the best of the authors’ knowledge, no studies exist comparing longitudinal strain and strain rate values using STE from different echocardiographic views in cats. Therefore, we examined the feasibility to obtain longitudinal strain and strain rate from the RP4Ch view in cats; moreover, we evaluated the agreement between RP4Ch and LAP4Ch views for assessment of LV longitudinal strain and strain rate and intra-observer variability.
We included 50 client-owned cats in the study: 31 clinically healthy cats and 19 with different disease states. The breeds of the cats included in the study consisted of domestic shorthair (47), Siamese (1), Persian (1) and Maine coon (1). Of the 50 cats, 20 were females and 30 males, with a median age of 60 months (range: 6 months–18 years) and weighing a median of 4 kg (range: 2–9 kg). The median heart rate recorded during the RP4Ch STE acquisition (191 beats/minute; range: 130–243 beats/minute) did not differ from that recorded during the LAP4Ch STE acquisition (187 beats/minute; range: 119–244 beats/minute; p = 0.66). Of the 19 diseased cats, 15 had cardiomyopathy and 4 had extracardiac disease. Of the cats with cardiomyopathy, 11 cats had hypertrophic cardiomyopathy, 2 cats had restrictive cardiomyopathy and 2 cats had unclassified cardiomyopathies. Extracardiac diseases included lungworm infestation (2 cats), chronic kidney disease (1 cat) and diabetes mellitus (1 cat).
We obtained the systolic peak of the LV longitudinal strain and strain rate values of endocardial and epicardial border from the RP4Ch and Lap4Ch views in all cats (Table 1
We found no fixed or proportional bias when comparing LV longitudinal strain and strain rate values of endocardial and epicardial border obtained from the two different echocardiographic views. Endocardial longitudinal strain and strain rate had 95% limits of agreement from −3.28 to 2.58 and −1.41 to 1.36, respectively (Figure 2
Epicardial longitudinal strain and strain rate had 95% limits of agreement from −11.58 to 9.19 and −2.28 to 1.74, respectively (Figure 3
Intra-observer measurement variability from RP4Ch view was <10% for LV longitudinal strain values and <20% for LV longitudinal strain rate values. Intra-observer measurement variability of all LV STE variables for both echocardiographic views is listed in Table 2
Our study demonstrates that longitudinal LV strain and strain rate values obtained from the RP4Ch or LAP4Ch view are similar in healthy and diseased cats. Indeed, the methods show no fixed or proportional bias, suggesting that they provide similar estimates of longitudinal LV deformation. However, only endocardial longitudinal strain values obtained with the two views agreed sufficiently to be used interchangeably. Conversely, the agreement between the two echocardiographic views for epicardial longitudinal strain/strain rate and endocardial longitudinal strain rate values was insufficient to recommend interchangeability. Intra-observer measurement variability from RP4Ch view was acceptable for all STE variables.
STE is an advanced imaging technique useful to assess myocardial function by strain and strain rate analysis. This novel echocardiographic modality is angle-independent, therefore beam angles do not influence the deformation analysis. Recently, the feasibility for assessment of the longitudinal strain and strain rate using STE from the RP4Ch view has been described in dogs [18
]. This approach allowed investigators to obtain a good quality four chamber image for STE analysis, but, as the authors reported, a suboptimal tracking of the apical segments, especially during the ventricular diastole, could cause an inadequate analysis [18
]. These authors concluded that data obtained from RP4Ch and LAP4Ch views were not considered interchangeable [18
]. Our findings differed from those of the previous investigators; indeed, apical segments of the LV were easily visualized, and we could track the points used to delineate the endocardial and epicardial borders. Moreover, the software followed the myocardium during the entire cardiac cycle and the LV apex did not go out of the sector scan. We speculate that this different with previous study in dogs may be explained by the cardiac dimensions and orientation of the heart within the chest in cats compared to the dogs. Commonly, the LV was entirely visualized in the sector scan by RP4Ch view during the entire cardiac cycle in cats. On the other hand, in dogs, especially large breeds, the LV apex can move out of the scan sector in RP4Ch view.
A complete echocardiographic examination includes the visualization of both RP4Ch and LAP4Ch views in cats [18
]. The LAP4Ch view is commonly used for LV Doppler and tissue Doppler imaging evaluations because an optimal alignment with blood flow or mitral annulus can be obtained. Other investigators have examined STE analysis of the LV from LAP4Ch view in cats [13
]. However, RP4Ch view is often easier to obtain than the LAP4Ch view and should be considered when LAP4Ch view is suboptimal. Our results showed that strain and strain rate values were similar between the two echocardiographic views, but only endocardial longitudinal strain values were interchangeable. Contrary to the previous study in dogs [18
], we could observe a difference between endocardial and epicardial deformation analysis because our software permitted separate evaluation of these components of the LV strain and strain rate.
Some authors have reported that RP4Ch view cannot include the true apex of the LV in dogs [25
]; however, similar problems occur in humans from the LAP4Ch view [26
]. Recently, several studies have reported that foreshortening of the LV in RP4Ch view was not evident if the view was imaged correctly in dogs [27
]. In our study, to avoid a false visualization of the LV apex and a shortened LV long axis, we took care to obtain a LV image in both views with the longest possible long-axis dimension.
STE analysis requires high echocardiographic image quality to follow the speckles frame-by-frame during the cardiac cycle. Our software easily and consistently followed the epicardial and endocardial borders during the cardiac cycle in both echocardiographic views. However, although epicardial deformation variables and endocardial strain rate showed similar values using different echocardiographic views, we did not consider them interchangeable, because we deemed the absolute differences or the percentage differences too large. This may suggest that tracking of the epicardial border is suboptimal compared to endocardial border, possibly because of interference from pericardium or lungs. Moreover, whether or not STE has a certain intrinsic angle-dependence which can influence strain and strain rate values in different views, remains undetermined.
Intra-observer measurement variability showed low values for strain analysis in both echocardiographic views, while higher intra-observer measurement variability was observed for strain rate analysis. These findings agree with our previous STE studies of left atrium and right ventricle [8
]; therefore, we cannot exclude that this high variability for some variables may be software dependent.
Our study has some limitations. We have evaluated only the global performance of the LV (average strain and strain rate of all segments, while in the previous study in dogs [18
], the investigators analyzed and compared individual myocardial segments. However, veterinary clinicians are mostly interested in assessing global strain and strain rate, rather than regional dyskinesis, because myocardial infarction occurs infrequently in dogs and cats. Furthermore, we performed the strain and strain rate analysis using XstrainTM
software, so our findings cannot be generalized to other STE software. Indeed, data obtained from software of different vendors are not interchangeable [31
]. Finally, in our study population no cats with dilated cardiomyopathy or end-stage hypertrophic cardiomyopathy were present. The dilated LV, typically observed in these cardiomyopathies’ forms, could not permit to visualize the LV apex within the sector scan from RP4Ch view.