Left atrial dynamics in the hemodynamic response during exercise in patients with asymptomatic severe aortic stenosis: a cardiopulmonary imaging study

Abstract

Abstract Background The hemodynamic impact of left atrial (LA) dynamics in aortic stenosis (AS) in relation to cardiopulmonary response to exercise has never been studied. We aimed at investigating the link between LA function vs valvulo-arterial impedance (Zva) and right ventricular (RV)-to-pulmonary circulation (PC) coupling in asymptomatic severe AS patients. Methods A total of 94 patients: 64 asymptomatic severe AS patients (aortic valve area (AVA) <1.0 cm2 or AVA index <0.6 cm2/m2) with ejection fraction >50% and 30 gender-matched control subjects underwent cardiopulmonary exercise testing combined with Echo-Doppler with assessment of LA strain and RV-to-PC coupling (tricuspid annular peak systolic excursion (TAPSE)/ pulmonary arterial systolic pressure (PASP) ratio). AS patients were divided into 3 groups according to peak aortic jet velocity (PV), mean pressure gradient (MPG) and stroke volume index (SVI). Zva was assessed using (MPG + systolic blood pressure)/ SVI ratio. Results Paradoxical low-flow low-gradient AS (PLFLG: PV <4 m/s and MPG <40 mmHg, SVI ≤35ml/m2, N=18, AVA 0.77±0.16 cm2), Normal-flow low-gradient AS (NFLG: PV <4 m/s and MPG <40 mmHg, SVI >35ml/m2, N=23, AVA 0.85±0.16 cm2) and High-gradient AS (HG: PV ≥4 m/s or MPG ≥40 mmHg, N=20, AVA 0.62±0.17 cm2) had a higher LA volume index than Control (Control 22±6, PLFLG 33±11*, NFLG 38±12* and HG 33±9* ml/m2, *P<0.05 vs Control). There was no significant difference in peak VO2 (17±5 ml/min/kg) and VE/VCO2 slope (28±3) among 3 AS groups although PLFLG had lower peak cardiac output (7.0±2.4 L/min) compared to NFLG (9.0±2.3 L/min) and HG (9.2±3.3 L/min). In PLFLG and NFLG AS, LA strain at rest (21±9 and 26±13%) and during exercise (26±12 and 31±14%) were decreased compared to Control (37±8% at rest, 43±11% during exercise) but maintained some reserve during exercise (P<0.001). HG AS had no increase in LA strain (31±15% at rest, 28±15% during exercise) (Figure A). In AS groups, no significant correlation at rest was observed between LA strain and Zva, whereas a negative correlation was observed during exercise (R=−0.4, P=0.003, Figure B). LA strain was also correlated with TAPSE/PASP at rest and exercise (R=0.44 and 0.47, P<0.01, respectively, Figure C). Conclusions In asymptomatic severe AS, the study of LA functional adaptation to exercise plays a key role in the hemodynamic unfavorable cascade from AS-related left ventricular afterload to RV-to-PC uncoupling. Funding Acknowledgement Type of funding source: None