Left atrial function index: did we end up waiting for Godot?

‘Estragon: What do we do now? Vladimir: Wait for Godot.’1 Example of left atrium 2D speckle tracking longitudinal strain in a normal subject. The three phases of atrial cycle are underlined with different colors. Please note that reservoir and conduit phases provide positive curves whereas contraction leads to a more or less negative one. Thus, ‘we were still waiting for …’ In their recent article, Sargento et al.8 proposed the use of a new parameter for the assessment of LA and tested its effectiveness as prognostic predictor in patients with HF and reduced EF (HFrEF). The so-called left atrial function index (LAFI) was for the first time proposed by Liza et al.9 in 2008 and it is calculated as LAFI = (LAEF × LVOT−VTI)/LAESVi, where LAEF is the LA emptying fraction, obtained by LA end-diastolic and end-systolic volumes, LVOT-VTI is the velocity-time integral across the LV outflow tract (LVOT) and LAESVi is the LA indexed end-systolic volume. The first clear advantage of this parameter so calculated is that it includes—in the same time—analogues of cardiac output, atrial reservoir function and LA size. It allows LAFI to provide both anatomical and functional information, relating it to the LV systolic function. The LAFI increases proportionally to LA reservoir function and stroke volume but is inversely proportional to LA size; it means that LAFI decreases in all the conditions that lead to an impairment of LA reservoir or LV systolic function and an increase of LA dimensions. Thus, LAFI appears to be a very sensitive but not very specific marker of LA and LV diastolic dysfunction. The second important advantage of this measurement is that it is rhythm independent.9 Indeed, it depends on LA emptying fraction that provide an estimation of atrial function, but is not tightly related to atrial contraction; therefore, it can be used in patients with atrial fibrillation. Sargento et al.8 confirmed these speculations. They analysed 203 outpatients with a LVEF <40%, who were clinically stable and on optimal therapy, measured LAFI and categorized patients into quartiles according to LAFI values. Moreover, they quantified all-cause death rate in their cohort of patients, built four different multivariable models including different potential confounders and compared the effectiveness of LAFI with these models and with each of its individual components (LAEF and LAESVi) for the prediction of survival. They found that LAFI was significantly correlated to several clinical and echocardiographic parameters, including NYHA, N-terminal pro brain-type natriuretic peptide, heart rhythm, therapeutics, LV global strain, and atrial strain. Moreover, they showed that survivors had higher LAFI values than deceased patients and that the death rate increased from the highest LAFI quartile to the lowest, with an optimal LAFI cut-off point to predict death ≤ 24.5 (sensitivity and specificity of 88.5% and 47.9%, respectively). More interestingly, LAFI was superior to each of its individual components in predicting the outcomes, resulting in the only significant predictor of survival in the direct comparison with them. It means that the association between the anatomical and functional features of LA and the inclusion of LV systolic function in the index provide a better prediction of survival compared with the individual elements. Accordingly, LAFI increased the prognostic effectiveness of the four clinical and echocardiographic multivariable models, showing that its use should be associated with the current measurements for a more tailored risk stratification of patients with HFrEF. Finally, the authors demonstrated that LAFI and increasing LAFI quartiles were univariate predictors of survival in patients with atrial fibrillation, overcoming the most significant limitation that previously had affected valid parameters like A and A’ velocities. On the other hand, the very high sensitivity of LAFI is partially counteracted by a low specificity, which makes a high LAFI a strong predictor of survival, rather than a low LAFI a strong predictor of death. However, most of the best echocardiographic parameters currently used in everyday routine, from LVEF to global longitudinal strain, are affected by a low specificity, providing fundamental information for the physicians at the same time. The prognostic value of LAFI had been previously underlined by Welles et al.,10 which found that LAFI predicted HF hospitalization in patients with coronary heart disease and EF ≥50%, with an increased risk of adverse cardiovascular outcomes from the higher to the lower LAFI quartiles. Of course, all these findings need to be confirmed in larger cohorts of patients and in different clinical settings, particularly HFpEF. However, the great results obtained so far and the easily understandable potential of a complete parameter let us hope that, maybe, we end up waiting for Godot. Conflict of interest: None declared.