Central role of left atrial dynamics in limiting exercise cardiac output increase and oxygen uptake in heart failure: insights by cardiopulmonary imaging

Aims In heart failure (HF), the haemodynamic response to exercise in relation to left atrial (LA) dynamics is unexplored. We sought to define whether abnormal LA dynamics during exercise may play a role in cardiac output (CO) limitation and tested its ability to predict outcome. Methods and results A total of 195 HF patients with reduced (n = 105), mid-range (n = 48), and preserved (n = 42) left ventricular ejection fraction (LVEF) and 46 non-cardiac dyspnoea (NCD) subjects underwent cardiopulmonary imaging with assessment of LA strain and strain rate (sra). HF patients, irrespective of LVEF, exhibited a significantly impaired LA strain and LA sra at rest, during exercise and recovery compared to NCD subjects with a blunted response in CO and right ventricular to pulmonary circulation coupling. LA strain and LA sra during exertion were significantly correlated with peak stroke volume index, peak CO and peak cardiac power output (R = 0.33, 0.48, 0.50 and R = 0.36, 0.51, 0.52 for LA strain and LA sra, respectively; all P < 0.001). In HF, after adjustment for age, gender, left atrial volume index, peak mitral regurgitation > 2, peak LVEF and peak heart rate, a higher LA strain (1% increase) during exercise was associated with a higher peak stroke volume index (mL/m(2)), CO (L/min) and cardiac power output (mmHg L/min) at multivariable analyses (beta-coefficients +/- standard error = 0.23 +/- 0.07, 0.046 +/- 0.014 and 0.012 +/- 0.004, respectively; P < 0.05). The same associations were found also for LA sra (absolute value) (1/s increase, beta-coefficients +/- standard error = 1.91 +/- 0.68, 0.43 +/- 0.14 and 0.12 +/- 0.04, respectively; P < 0.05). Exercise LA strain (adjusted hazard ratio 0.94, 95% confidence interval 0.92-0.97, P < 0.001) and LA sra (adjusted hazard ratio 0.60, 95% confidence interval 0.43-0.84, P = 0.003) were associated with a worse outcome after adjusting for age and gender. Conclusions In HF, LA dynamics may play a key role in exercise CO increase due to an impaired forward (left ventricular filling) and backward (right ventricular to pulmonary circulation uncoupling) combination of mechanisms, irrespective of LVEF. The blunted LA strain and LA sra reserve during exercise are associated with a worse cardiopulmonary performance and outcome.