Left atrial volume (LAV) emerged as a marker of cardiovascular disease many years
ago with a large body of supporting scientific evidence. Since then, it has found
clinical utility in the cardiovascular field, especially in cardiology and neurology.
It remains of great interest as to why so simple a measure carries such a strong prognostic
impact for a wide variety of cardiovascular diseases [1,2]. Among these disorders
are ischemic heart disease, heart failure (HF) [[3], [4], [5], [6]], hypertrophic
cardiomyopathy (HCM) [7], atrial fibrillation (AFib) and even for predicting stroke
in the absence of AFib. To understand why LAV is so sensitive [8], we look to the
mechanisms of left atrial remodeling with a common endpoint of increased LAV, left
atrium (LA) dilatation and hypocontractility, and eventually resulting in AFib [1]
(Table 1). Identifying increased LAV will allow early detection of LA dysfunction
[1] and alert clinicians to treat the patient in a timely manner to reduce the negative
consequences of long-standing incompletely treated diseases or pathologic states (Table
1).
Table 1
Underlying pathophysiologic mechanisms that affect LA size and function.
Table 1
Pathophysiologic mechanism
Effect on LA
Examples of diseases and pathologic states
Pressure and/or volume overload
Hypertrophy, necrosis, apoptosis and fibrosis at cellular and extracellular matrix
levels [1]
Neurohumoral activation such as atrial natriuretic peptide (ANP) [1,18]
•
LV systolic dysfunction (HFrEF) [3,4]
•
LV diastolic dysfunction (HEpEF) [5,6]
•
Ischemic heart disease [3,4]
•
Post-myocardial infarction [3,4]
•
Hypertension
•
Aortic stenosis
•
Aortic insufficiency
•
HCM [7]
•
Mitral stenosis
•
Mitral regurgitation
•
LV dilatation [3,4]
•
LV myocardial scar [3,4]
•
Iron deposition on the myocardium
•
Sarcoidosis
•
Amyloidosis
•
Obesity
•
LV hypertrophy on athletes
•
Aging
•
Dilated cardiomyopathies
•
Obstructive Sleep Apnea
Tachycardia
•
AFib
•
Hyperthyroidism
•
Drug induced tachycardia
•
Energetic beverages
•
Caffeine in excess
Abbreviations: HFrEF: Heart failure with reduced ejection fraction, HEpEF: Heart failure
with preserve ejection fraction, HCM: Hypertrophic cardiomyopathy, LV: Left ventricle,
AFib: Atrial fibrillation.
The power of LAV has been tested in long-term follow up of patients, underscoring
its predictive nature [9]. This predictive capability has been largely validated with
echocardiography (Echo), where its value was originally confirmed, and with other
newer and more accurate methods such as cardiovascular magnetic resonance (CMR) and
contrast-enhanced cardiac computed tomography (CT). The latter has traditionally been
performed using retrospective acquisition of coronary computed tomography angiography
(CCTA) images in different pathologic states (such as those described in Table 1).
The accuracy of the different imaging modalities has been largely documented, however
two-dimensional echocardiographic (2D-echo) methods are somewhat limited, [10] three-dimensional
(3D) echocardiography (3D-echo) has better accuracy and reproducibility but is not
widely used in routine clinical practice (despite newer rapid imaging and analysis
strategies that improve its utilization) [10], atrial strain imaging can detect LA
dysfunction before the manifestation of LA structural changes since its parameters
are influenced by loading conditions, rhythm irregularity and fibrosis [11], but,
as it is for 3D-echo, is not widely used in routine clinical practice. CMR is the
gold standard method for the assessment of cardiac structure and function [12] due
to its strong accuracy and reproducibility, and LAV reference ranges are published
and validated for different gender and ages [13]. CCTA, based on its retrospective
acquisition, allows 3D visualization of the LA and is readily available for analysis
without additional testing with high spatial resolution across the cardiac cycle [14].
In the real world, obtaining this strong predictor in asymptomatic population is limited,
since it is not feasible to perform 3D-echo, CMR and CCTA just to obtain the LAV in
this group. 3D-echo and CMR are not ubiquitous, and CCTA requires somewhat greater
radiation doses, even with dose-limiting protocols [15], since it necessitates retrospective
acquisition and the use of contrast media.
Thus, given the clinical relevance of LAV assessment, it would be preferable to have
a highly accurate and reproducible method such as those described, but also one that
is widely used and that overcomes the real-world limitations described.
Toward this end, Cardona et al. developed a method to test whether a highly reproducible
and accurate LAV could be obtained on more commonly-performed non-contrast coronary
artery calcium (CAC) scans. Their study is particularly clinically relevant since
CAC scans are widely available, have a well-established operator-independent protocol,
do not require contrast media, and are acquired with radiation doses significantly
lower than CCTA.
Cardona et al. in their recent publication [16] provide a valuable contribution by
measuring LAV using the standard CAC protocol, and proving that it is feasible and
highly accurate compared with the reference standard of CCTA. They overcame certain
technical limitations that are attributed to non-contrast CT studies, including not
having to modify the standard acquisition protocol, and developing a method to use
the mitral valve plane as a landmark in a multiplanar reformatting approach. This
new method reduced an important source of variability and low reproducibility, and
provided excellent interobserver and intra-observer reproducibility. The authors identified
several additional technical characteristics of the non-contrast gated CT scan and
of the LA anatomy that allows this approach to be feasible. These included the fact
that both CAC scan and CCTA are ECG-gated, that the LA is surrounded by easily identifiable
structures on non-contrast CT images, the lower density of the interatrial septum,
and that the mitral annulus has specific and easily identifiable landmarks (fat in
the auriculo-ventricular groove, the circumflex artery and the coronary sinus).
There is a difference between LAV from CAC scan and from the reference standard on
CCTA, due to technical reasons such the potential inclusion of part of the esophageal
wall, the overestimation related to the thicker slices of the CAC scans, and the decreased
endocardial border definition of the non-contrast enhanced study. Even though these
differences were statistically significant, they had little clinical relevance since
they only represent 4% of the average measured LA volume.
Normal reference values will need to be established for this approach in an asymptomatic
population since the reported values are measured at end-systole and not in mid-diastole
as for CAC scans. Likewise, the technique will require validation in symptomatic individuals
with pathologic states (Table 1), and in long-term follow-up cohorts to validate its
predictive strength.
Cardona et al. have advanced the field by proving that standard CAC scans can be used
to quantify LAV. This tool allows the benefit of LAV measurements to be applied to
routine CAC scans, and should help to improve cardiovascular risk prediction in asymptomatic
populations mainly for those subjects with new-onset AFib [17]. It is likely that
using this method will reclassify the cardiovascular risk of an individual, above
and beyond the metric of the CAC score and in addition to conventional risk stratification
scales.
Conflict of interest
The author declare that there is no conflict of interest.