The left atrium: a mirror of ventricular systolic and diastolic function

This editorial refers to ‘Potential usefulness and clinical relevance of a novel left atrial filling index to estimate left ventricular filling pressures in patients with preserved left ventricular ejection fraction’, by K. Braunauer et al., pp. 260–269. Braunauer et al.1 investigate a novel index of left ventricular (LV) diastolic function which they name the left atrial (LA) filling index, calculated as the ratio between peak early transmitral flow velocity (E) and LA reservoir strain. They studied the clinical usefulness of the index in patients with LV preserved ejection fraction with either heart disease or cardiovascular risk factors, and showed in a small subgroup that the E/LA strain ratio was associated with LV end-diastolic pressure, and suggest that this index may be used to estimate LV filling pressure. Furthermore, in a cohort of 486 patients they showed that high values of E/LA strain were associated with aggravation of heart failure after 2 years. Using speckle tracking echocardiography, measurement of LA strain is feasible in most patients. LA reservoir strain is measured from LV end diastole to peak positive LA strain. It is recommended to measure LA strain from an optimized (i.e. non-foreshortened) apical four-chamber view of the LA.2 Biplane LA strain from apical four- and two-chamber views is another option. Strain is calculated as the average value from all LA segments. LA reservoir strain is related to LA pressure and was proposed as a non-invasive marker of LV filling pressure.3 Low values for LA reservoir strain are associated with elevated LV filling pressure. Average values for normal LA strain were reported as 42% in the NORRE study4 and as 46% in a study by Morris et al.,5 and in these studies the lowest expected value of LA reservoir strain was reported as 26% and 23%, respectively. Which cut-off value for LA reservoir strain to use as marker of elevated LV filling pressure has so far not been well defined? There is age-dependency as LA reservoir strain decreases with age along with reduction in mitral E-velocity.6 Toma et al.,7 in a clinical study suggested that LA contractile function also helps to facilitate LA early filling (Figure 1). Consistent with this, Barbier et al.,8 in an experimental study showed that LA reservoir filling has an early phase which is due to atrial relaxation and a late phase which they attributed to LV longitudinal shortening causing systolic descent of the cardiac base. As shown by wave intensity analysis, the early LA reservoir phase represents atrial filling by suction due to LA contraction and relaxation, whereas the late phase is due in part to suction caused by systolic descent of the cardiac base, but is also attributed to the right ventricular pressure pulse which ‘pushes’ blood into the LA.9 Early and late reservoir strain may be differentiated when pressure is measured along with strain.10 As shown by Barbier et al.,8 increase in atrial stiffness decreases LA late reservoir function. Left panel: Recording from a patient with normal LV function, showing simultaneous LA pressure and volume. Modified from Toma et al.7 The A, C, and V waves and X and Y descents of LA pressure are indicated; x, point of nadir pressure of X descent; a, onset of atrial ejection phase; min, point of minimum LA volume; max, point of maximum LA volume; Va, Vmin, Vx, and Vmax, LA volumes at a, min, x, and max points, respectively. Please note that LA reservoir volume has two phases, an early phase between Vmin and Vx, when LA pressure is falling and a late phase between Vx and Vmax when LA pressure is rising. Right panel: Determinants of LA reservoir strain. The schematics of Figure 2 is from a clinical study which investigated determinants of LA reservoir strain, and concluded that the strongest determinant was LV global longitudinal strain.11 Furthermore, LA volume and LA pressure estimated as pulmonary capillary wedge pressure were independent determinants of LA strain. In that study, the contribution of LA relaxation was not addressed. Importantly, the interaction between LA strain and LV global longitudinal strain implies that LA strain is determined not only by LV diastolic function but also to a large degree by LV systolic function. Schematic illustration of modifiers of LA reservoir strain. LV systolic function - panel A: Reduction in LV longitudinal shortening reduces atrial stretch and reservoir strain. LA volume - panel B: Enlargement of the atrium leads to smaller LA strain for a given reservoir volume. LA pressure - panel C: Elevation of LA pressure implies stiffening of the atrium and therefore lower strain for a given rise in pressure. The figure is based on Andersen et al.11 The rationale for using E as the numerator of the E/LA strain ratio is that a high value for E, implies a large transmitral pressure difference, which in most cases reflects elevated LA pressure. Calculation of peak pressure gradient is not feasible from mitral E since a substantial fraction of the transmitral pressure gradient is needed to overcome inertia. Therefore, the real pressure gradient is larger than predicted by the simplified Bernoulli equation. However, the peak transmitral filling rate correlates very well with the transmitral pressure gradient.12 Since both a high E and a low LA strain are associated with elevated LV filling pressure, the E/LA strain ratio increases progressively with increasing LV filling pressure. Normal LA strain indicates a compliant atrium, which by pooling pulmonary venous return flow during systole and supplying it to the ventricle during diastole, facilitates ventricular performance.13 Therefore, LA reservoir strain is important for overall cardiac performance. In summary, the approach proposed by Braunauer is interesting since it incorporates two important variables related to LV diastolic function in one index. This is also a limitation since it is not clear if mitral E and LA strain always behave consistently. Studies with more extensive validation of the relationship of E/LA strain ratio to invasively measured LV filling pressure are needed, and confounders of the relationship should be defined before we can be confident that this new index provides more accurate information about filling pressure than the established alternative measures. Furthermore, the clinical value of E/LA strain as a risk marker needs to be tested in prospective studies with relevant clinical endpoints. Conflict of interest: none declared. The opinions expressed in this article are not necessarily those of the Editors of EHJCI, the European Heart Rhythm Association or the European Society of Cardiology.