Trouble with estimating filling pressure in acute heart failure: lessons from Takotsubo syndrome

This editorial refers to ‘Left atrial strain analysis improves left ventricular filling pressures non-invasive estimation in the acute phase of Takotsubo syndrome’, by G. Iannaccone et al., https://doi.org/10.1093/ehjci/jead045.

In chronic heart failure, left ventricular (LV) filling pressure is estimated with good accuracy by applying a combination of echocardiographic parameters.^1–3 In acute patients, however, these parameters are less useful for estimation of LV filling pressure.^4,5 Iannaccone et al.⁶ tested these parameters in Takotsubo syndrome. In addition, they tested the novel parameter left atrial (LA) strain as a marker of LV filling pressure. The study included 62 patients in the acute phase of Takotsubo syndrome. The patients had markedly elevated LV end-diastolic pressure (LVEDP) and moderate reductions of LV ejection fraction and global longitudinal strain.

Takotsubo syndrome is associated with a high risk of major complications. In the study of Iannaccone et al., the most common in-hospital complication was a composite of pulmonary oedema and/or cardiogenic shock or Killip Class III/IV, which occurred in patients with a mean LVEDP of 30 mmHg and mean LV pre-atrial contraction pressure (reflecting mean LA pressure) of 20 mmHg. These observations are consistent with the notion that excessive LV filling pressure causing pulmonary congestion is a driver for complications in Takotsubo syndrome. Therefore, accurate non-invasive methods for assessment of LV filling pressure are important for estimation of risk in patients with Takotsubo syndrome.

The most important novel finding in the study of Iannaccone et al. is demonstration in acute heart failure patients that LA reservoir and pump strains were superior to the conventional echocardiographic markers of LV filling pressure. Left atrial reservoir and pump strain decreased progressively with elevation of LVEDP. Both LA strain parameters had strong correlations with LVEDP, with r-values of −0.859 and −0.848, respectively. These are better correlations between LA strain and LV filling pressure than reported in chronic heart failure patients.^7,8 Importantly, LA volume index (LAVI) had only a weak association with LVEDP, and peak tricuspid regurgitation velocity (TR velocity) did not correlate with LVEDP. The ratio between peak early mitral flow velocity and the average of septal and lateral mitral annular velocity (E/e′), however, showed correlation with LV filling pressure comparable to observations in chronic patients^2,3 and in a previous study of Takotsubo syndrome.⁹

The area under the receiver operating characteristics curve (AUCs) for the ability to predict LVEDP above the mean (>24.5 mmHg) were 0.90 and 0.89 for reservoir strain and pump strain, respectively. For E/e′, the AUC was 0.80 and not significantly different from AUCs for LA reservoir and pump strain. The AUCs for LAVI and peak TR velocity as predictors of filling pressure were not statistically significant.

The study by Iannaccone et al. is limited by its relatively small size. Furthermore, the study included only patients with an established Takotsubo diagnosis. This is different from clinical practice where acute patients may have a wide range of different diagnosis, including non-cardiac disorders. The promising observations of LA strain as a parameter of LV filling pressure in Takotsubo syndrome should be further explored in studies that include acute patients with different cardiac diseases. In addition, it would be interesting to assess long-term risk. As shown in a study that measured atrial strain by cardiovascular magnetic resonance feature tracking, disturbed LA strain in the acute phase of Takotsubo was a predictor of long-term mortality.¹⁰

Left atrial volume is an established marker of LV filling pressure in chronic heart disease. In acute patients, however, the diagnostic role of LAVI remains to be determined. In previous studies in acute coronary syndrome, there were rather weak associations between LAVI and LV filling pressure, and LAVI was often in the normal range even in patients with elevated filling pressure.^4,5 This is consistent with the study of Iannaccone et al. in acute Takutsubo syndrome, which showed LAVI of 34 ± 12 mL/m² in the total population, and in a subgroup with LVEDP of 30 ± 7 mmHg, there was only a mild elevation of LAVI to 39 ± 13 mL/m².

Enlargement of the LA in chronic heart failure is the result of a remodelling process that takes days and weeks. In acute patients with short histories, there is not enough time for substantial atrial remodelling, and therefore, LAVI is often normal. In contrast, LA reservoir strain drops immediately when there is LV systolic dysfunction with a reduction in longitudinal shortening. This is because the atrium is tethered to the ventricle and reservoir strain is determined predominantly by LV longitudinal shortening.^8,11 For pump strain, there is also an immediate reduction in response to LV dysfunction, mainly due to elevated LVEDP that represents LA afterload. Due to these atrio-ventricular interactions, LA strain gives instant information about LV systolic and diastolic function. There is also an increase in LA volume in response to an increase in LA pressure, but not to the extent seen in chronic heart failure where LAVI is an important diagnostic marker.

It is not clear why TR velocity did not correlate with LVEDP in the study by Iannaccone et al. A relatively large fraction of patients with Takotsubo syndrome have combined right and left ventricular dysfunction.^12,13 Since there was no presentation of data on estimates of right atrial pressure, TR velocity may have underestimated pulmonary artery pressure in some of the patients.

Since severe complications in Takotsubo syndrome are caused by pulmonary congestion, a non-invasive index like LA strain that can identify patients with excessive LV diastolic pressures can be important for patient management.

Conclusion

As shown by Iannaccone et al. in the acute phase of Takotsubo syndrome, LA reservoir and pump strain are excellent markers of elevated LV filling pressure. Consistent with previous studies in acute patients, LAVI was of limited value in the acute setting. Due to slow response to elevated LV filling pressure, LAVI is useful predominantly in chronic conditions. Left atrial strain that responds instantly to changes in LV filling pressure is well suited as a marker of filling pressure in acute as well as in chronic patients. Larger prospective studies are needed to further explore the role of LA strain as a marker of filling pressure and risk in Takotsubo syndrome and in other acute cardiac disorders.

References

Author notes