Background Left atrial (LA) compliance and contractility are key determinants of left ventricular (LV) stroke volume. LA remodelling and dysfunction in heart failure with preserved ejection fraction (HFpEF) are considered to be late presenting features, secondary to hypertension-induced LV diastolic dysfunction and chronic LA pressure overload. However, a recent unifying hypothesis has stipulated that systemic and coronary microvascular inflammation drive HFpEF pathophysiology. Considering this paradigm, we hypothesised that diminished LA compliance and contractile function occur early during HFpEF development and further impair cardiac performance.

Methods Cardiac magnetic resonance imaging, echocardiography, invasive LV and LA pressure volume (PV) analysis and tissue analysis were performed in an experimental model of early-stage HFpEF (elderly dogs with renal wrap induced hypertension and exogenous aldosterone; n = 9) and young control dogs (sham surgery; n = 13).

Results Early HFpEF was associated with LA enlargement, LA myocyte hypertrophy, and enhanced LA contractile function (LA active emptying fraction 18 ± 6 vs 12 ± 2%, p = 0.03; LA end-systolic PV relationship slope: 2.7 ± 0.9 HFpEF vs 1.5 ± 0.6 mmHg/mL controls, p = 0.01). However left atrioventricular systolic coupling was impaired (ratio of LV to LA end-systolic PV relationship slope 1.3 ± 0.4 HFpEF vs 0.8 ± 0.4 Controls, p < 0.05) and the curvilinear end-LA reservoir PV relationship was shifted upward/leftward in HFpEF (LA chamber stiffness constant, ß, 0.08 vs 0.16 mmHg/mL in controls, p = 0.005) indicating reduced LA compliance. Impaired atrioventricular coupling and lower LA compliance were associated with lower LV stroke volume. Total LA and LV fibrosis were similar between HFpEF and controls although peri-pulmonary vein fibrosis was selectively increased in HFpEF. LA microvascular density tended to be lower in HFpEF than controls (2154 ± 578 vs 2650 ± 523 vessels/mm², p = 0.07) and inversely correlated with LA ß (r=-0.66, p = 0.006). Titin isoform composition (N2BA/N2B ratio) was similar between groups, but titin hyperphosphorylation was seen in HFpEF LA and LV tissue. LA titin hyperphosphorylation correlated with LA ß (r = 0.54, p = 0.03). LA microvascular reactivity was also diminished in HFpEF versus controls.

Conclusions Maladaptive LA remodelling is present in early-stage experimental HFpEF and compromises overall cardiac performance. Reduced LA compliance in HFpEF may relate to contributions from LA ischaemia and titin-based cardiomyocyte stiffening. Thus, pathophysiological mechanisms involved in the development of HFpEF directly affect both atrial and ventricular myocardium.