Strain Echocardiography: Values And Limitations – My Experience

“If echocardiographers are to stand still, depend on standard 2D echo imaging using equipment produced a decade ago and not upgraded since, perform ‘‘ejectionfractionograms’’, focus primarily on the left ventricle and simply ‘‘eyeball’’ the other chambers, and avoid new methods such as strain imaging and contrast echo because they are perceived as ‘‘a waste of time’’, then I fear that echocardiography will be passed by. As the dinosaurs illustrated, we need to adapt and continue to evolve, or face the consequences”.
Alan Pearlman- JASE editor, 2010

My research experience focused on heart failure and I conducted medical research on the derivation and validation of novel echocardiographic approaches to myocardial deformation. I have been heavily involved with all projects using strain echocardiography at Duke University (~ 40 projects over the last 5 years). My research has focused on 1) using echocardiographic methods to understand cardiac diseases through in depth pheno-typing, and 2) Left and right atrium strain. I have completed > 10,000 speckle tracking strain measurements analysis on different cardiac diseases and on different cardiac chambers.

Speckle Tracking Strain: Speckle tracking strain imaging has been around for quite a while and we celebrated the first decade 3 years ago, whereas the technique described in 2004 (Leitman M-JASE) and clinical applications appeared around 2005 (Notomi Y- JACC). Since then, the interest has risen dramatically and so far we have > 5000 publications on this topic. Left ventricular (LV) ejection fraction (LVEF), the most widely used measure of cardiac function, has important limitations including low sensitivity for incident HF, technique-related variability and does not directly assess LV contractility. Global longitudinal strain (GLS) is the most studied among strain parameters and its prognostic value has been demonstrated in several clinical scenarios. GLS was widely adapted by cardio-oncology and some studies showed that less negative GLS values at completion of chemotherapy predicted future LV dysfunction, despite normal LVEF. GLS played a vital role in the screening, diagnosis, and management of hypertrophic cardiomyopathy, amyloidosis, valvular heart disease, heart failure and many other diseases where the reduction on LVEF may be a late phenomenon or less sensitive.

Inter-Vendor Variety: GLS inter-Vendor variety has dramatically dropped over a short period of time. However, still if we look at the same images with different software vendors, we may have different values. This difference is due to several factors: 1) Image quality and acquisition; 2) Software used; 3) Where to measure (endocardial, myocardial, median); 4) Post-processing of data; 5) Patients age, gender variabilities and loading conditions. Recent study by (Farsalinos KE- JASE, 2015) showed that the mean values of GLS between different software vendors was almost 3.6%. I still think this is a big difference, however the good news from this study is that the reproducibility of GLS measurement seemed to be very good and better compared to routinely used LVEF. The study also showed that GLS has much less variation than other traditional echo parameters (E/A, wall thickness, LVEDD, EF, and E’).

Regional Strain: With regard to regional (segmental) function, we now have, for the first time, the possibility to measure regional myocardial function. However, I think this is more challenging than GLS, and there is still progress to be made. The challenges become because we have only one segment to work with, and less data to average. Tracking quality becomes more important, regional artifacts matters more, definition of sample position more relevant, and more importantly, we cannot simply measure peak values anymore. For regional analysis, the strain curve shape (not peak values, because peak values can be the same), become critical. Regional strain measurements have much higher variability among vendors when compared with GLS. Furthermore, I have notice a big difference when I measuring the normal segment and the abnormal (scar) segment.

Strain Patterns: Specific strain patterns, such as that attributed to cardiac amyloid, I believe the severity of the disease matter more than the etiology. I think some of chronic global heart diseases states such as hypertension, aortic stenosis, heart failure with preserved EF, etc. share similar strain patterns and the pattern depends not on the etiology, but on the severity of disease states. For example, in more severe (late stage) aortic stenosis, or hypertension, I have notice similar strain patterns that has been described in cardiac amyloid, that affect mainly the basal/med segments and spares the apex. The LV apex is comprised mainly of circumferential muscle fibers, that are facing different hemodynamics load (BP) or may be histologically different from the longitudinal muscle fibers (basal/ med). On the other hand, do these chronic diseases also have accumulation of proteins similar to amyloid and not necessarily represent a complete picture of cardiac amyloid. Still an area of research!

Normal Values: What is the reference values? is another common question. Personally, I think no normal values could be given and each echo-lab should build up their normative values according to the local data, software vendor used, image acquisition, experience and the echo machine setup. Much remains to be done on improving the software, acquisition layered strain, and updating the myocardial function documents. For this discussion, I focused on the 2D LV GLS, though there is also promising research on the left atrium and right heart, and right atrium strain which have traditionally been hard to characterize.

In summary, I think GLS is ready for clinical practice. Its robust, reproducible and has been shown to add unique data that can guide diagnosis and management. I recommend GLS as a valuable complement to traditional function parameters. Further studies are needed to standardize vendors, recognizing specific strain patterns and to determine if there are age, gender variabilities or loading conditions difference.

Fawaz Alenezi Headshot
Dr. Fawaz Abdulaziz M Alenezi is a post-doctorate associate at the Duke University Health Systems. He conducts medical research on the derivation and validation of novel echocardiographic approaches to myocardial deformation and a new echocardiographic technique which assists patients with heart ventricular function.