The left atrium is an anatomically complex structure that has evolved to receive blood from the pulmonary veins efficiently and to transport it across the mitral valve to the left ventricle, a process that involves both active and passive components. Left atrial (LA) anatomy has been reviewed by Wang et al. and Ho et al. The left atrium consists of a venous component adjacent to the pulmonary vein orifices, the appendage, the vestibule of the atrioventricular valve, and the intra-atrial septum and free wall. Although the LA free wall is relatively thin (3.5–5.0 mm), its muscle fibers are arranged in two layers. One layer is superficial and circumferentially oriented and mainly occupies the vestibular region above the mitral annulus. It contributes to mitral valve competence and to the decrease in the anterior-posterior shortening recorded by parasternal imaging. The second set of fibers is subendocardial and longitudinal in orientation and extends from the LA roof to the mitral annulus. Contraction of these fibers results in shortening of the LA long axis and also elevation of the mitral annulus during atrial systole. Longitudinal strain (LS) and strain rate imaging measures the cyclic deformation resulting from longitudinal fiber contraction and relaxation. The left atrium also contains specialized muscle fibers (e.g., the Bachmann’s bundle) that facilitate electrical impulse conduction. The LA muscle fibers are embedded in an extracellular matrix containing collagen types I and III and elastin fibers, metalloproteinases, other proteolytic enzymes, and glycosaminoglycans. The intracellular matrix contributes to structural and electrical integrity of the left atrium, and when combined with its structural and geometric heterogeneity and the effects of surrounding cardiovascular structures (e.g., the right atrium, pulmonary veins, and aorta) contributes to the asymmetric nature of remodeling that may be induced by a variety of physiologic and pathologic conditions.
LA modulation of left ventricular (LV) filling is the result of its intrinsic cyclical deformation but also reflects alterations in cardiac electrophysiology and interactions with pulmonary vein flow (preload), mitral valve anatomy and performance, LV structure and performance (atrioventricular coupling), and the influence of aortic structural and hemodynamic alterations (arterioventricular coupling) that become increasingly more important with age. The left atrium modulates LV filling by four phasic mechanical functions: (1) atrial mechanical systole (active contractile pump) followed by (2) atrial myocardial relaxation that generates suction to initiate atrial filling from the pulmonary veins, (3) a reservoir function whereby inflow from the pulmonary veins initiated by atrial myocardial relaxation is coupled to LV contraction by apical motion of the mitral annulus, and (4) a conduit function that is initiated in early diastole by LV relaxation–induced suction and recoil of the mitral annulus toward the roof of the left atrium.
LA anatomy and performance can be assessed by Doppler echocardiography by using and integrating the results of a number of techniques: volumetric, Doppler, Doppler tissue, and strain and strain rate imaging. LA dimension and volume measurement recommendations were recently published by both the American Society of Echocardiography (ASE) and the European Association of Cardiovascular Imaging. Although the single anteroposterior LA diameter measured in the parasternal long-axis view can be rapidly performed and is reproducible, it may not accurately reflect the effects of atrial remodeling, partially because of constraints imposed by the posterior aortic wall. Measurement of LA volumes by two-dimensional echocardiography using the biplane multiple disk summation technique incorporates fewer geometric assumptions and is recommended by the ASE and the European Association of Cardiovascular Imaging as the preferred method to measure LA volume in clinical practice. Its major disadvantage is that this method is time consuming and requires apical views recorded to minimize LA foreshortening. Although three-dimensional echocardiography may eventually supersede two-dimensional echocardiography because of its superior accuracy, it is currently a research technique. A limitation of volumetric measurements is that usually LA volume is calculated at only one time point, whereas determination of LA performance requires measurements of LA volumes at three or more points in the cardiac cycle: maximum, minimum, and before the onset of the P wave.
Doppler echocardiographic imaging of LA deformation is reported to be able to detect preclinical abnormalities in LA performance. It can be performed by using either tissue Doppler or speckle-tracking echocardiography. The latter has a number of advantages over the former, chief among them its relative independence from incident angle. LA global and regional LS and strain rate measurements by speckle-tracking echocardiography require apical acquisition optimized to minimize foreshortening. However, LA strain imaging quality and its analysis reliability may be diminished compared with LV strain because of anatomic and physiologic differences between the left atrium and left ventricle. Thinner walls combined with greater regional temporal and spatial deformation heterogeneity (a result of its anisotropic macro- and microarchitectural properties), even in the absence of physiologic or pathologic remodeling, make measurements of two-dimensional LA global and regional LS difficult to quantify and standardize. Additionally, these abnormalities are amplified by atrial rhythm and conduction abnormalities. Image quality is also degraded by higher signal noise from surrounding structures and the location of the left atrium in the far field of apical images acquired using transthoracic echocardiography. Finally, image analysis packages, which currently are validated only for the left ventricle, may require optimization and validation for LA measurements.
Despite a number of publications that report potentially important clinical applications for two-dimensional echocardiographic LA deformation imaging, a lack of standardization in image acquisition and analysis protocols and terminology, and small sample sizes, are major impediments to establishing robust, reproducible normal values for LA deformation mechanics measurements, an essential requirement for widespread adoption of this technique for clinical use. Variations in acquisition protocols include the number of imaging planes (apical four chamber, two chamber, and long axis) and regional segments analyzed to construct global LS values. Although some investigators have used only the apical four-chamber view, others have used both the apical four- and two-chamber views, and some have added the apical long-axis view. Many reports subdivide the LA walls into six segments for each view to calculate global LS and strain rate, while others exclude some of the segments, especially those that incorporate the pulmonary veins (LA roof), LA appendage, interatrial septum, and those adjacent to the posterior aortic wall, from analysis because of poor tracking by the software.
Most important is the zero reference point for the initiation of strain imaging curves. Two reference points have been used: the QRS complex and the P wave. The first can be used for patients in atrial fibrillation and other atrial tachyarrhythmias or bradyarrhythmias (e.g., junctional rhythm), in addition to sinus rhythm. It has the advantage of ease of recording of the peak LS associated with atrial reservoir function. Use of the P-wave zero reference point allows recording of all of the phases of LA mechanical deformation in a physiologic manner. To et al. identified the first negative deflection after the P wave as “active atrial contraction,” the combined atrial relaxation (suction force) and reservoir phases as “atrial filling,” and atrial conduit function as “passive atrial emptying.” However, the last term may be a misnomer, as this phase is initiated by LV diastolic suction. The corresponding LA strain measurements can be abbreviated as: LSneg, LStot, and LSpos.
Three articles in the present issue of JASE address important issues related to LA deformation imaging that require clarification for LA LS to achieve clinical utility. Miglioranza et al. report reference ranges for LA LS and its phasic components in normal nonobese and nonathletic individuals, ranging from 18 to 75 years of age (61% women). LA LS was determined from biplane apical LA images subdivided into a total of 12 segments and then averaged using the peak P wave as the zero reference point in accordance with the recommendations (unpublished) of the European Association of Cardiovascular Imaging, ASE, and industry task force for the standardization of deformation imaging held in Seattle during the 2016 ASE Scientific Sessions. Miglioranza et al. analyzed the effects of age and gender on reference ranges, and the correlations of each of the LA LS phasic parameters with indices of LA and LV systolic and diastolic performance derived from three-dimensional echocardiographic images were analyzed. They report that analyzable biplane images could be recorded in 171 (70% of the total cohort) versus 230 (94%) for the four-chamber view alone, a difference they report as due to poor tracking of the LA roof, segments close to the mitral annulus, and the appendage. The normal ranges for LA LSpos, LStot, and LSneg are displayed in Table 1 and compared with those reported by Saraiva et al. The main differences between the two results are the higher values reported for both LSpos and LStot by essentially the same amount, suggesting that the differences in the number of LA views and segments averaged could have affected the measurement of LSpos measurements. Of note, although the LA LS measurements calculated only from the four-chamber view were tabulated in Supplemental Table 1 in Miglioranza et al , the number of patients reported was higher than for the biplane measurements. Only LSpos values were comparable; the other two measurements derived from the apical four-chamber view were lower. The lower value for LStot is concordant with the finding of Cameli et al. that peak LA LS was higher in the two- versus four-chamber view. However, because of the potential impact on clinical practice, a direct comparison of phasic LA strain measurements obtained from only the apical four-chamber view to those obtained using biplane acquisition is needed.
