Abstract
BACKGROUND: Left ventricular diastolic dysfunction (LVDD) is highly prevalent in hypertension (HTN) and is strongly associated with heart failure with preserved ejection fraction (HFpEF). However, conventional cardiac magnetic resonance (CMR) protocols lack guideline-recommended hemodynamic parameters for diastolic function assessment, thereby limiting their utility in LVDD evaluation. This study employed CMR feature tracking (CMR-FT) to analyze left atrial (LA) function in patients with subclinical HTN. We aimed to validate CMR-FT-derived LA strain as a marker for detecting LVDD with elevated left atrial pressure (LAP), thereby providing CMR parameters for diastolic dysfunction evaluation without additional scanning. METHODS: This retrospective study included 152 HTN patients and 60 healthy controls who underwent echocardiography and CMR. According to guidelines, patients with HTN were stratified into the HTN with normal LAP group (no LVDD or grade I LVDD) and the HTN with elevated LAP group (grade II or higher LVDD). LA strain parameters reflecting phasic functions (reservoir, conduit, and pump) were quantified with CMR-FT. Intergroup differences in left ventricular (LV) functional parameters and LA strain were compared using ANOVA, Kruskal-Wallis, or Chi-squared tests. Independent correlations of LA strain were identified by multivariable linear regression. The diagnostic performance of LA parameters for elevated LAP was evaluated by generating receiver operating characteristic (ROC) curves. RESULTS: Compared to healthy controls, both HTN (normal LAP) and HTN (elevated LAP) groups exhibited significant impairment in LA reservoir strain (εs: 33.90%±5.51% and 24.75%±4.91% vs. 42.29%±6.21%; P<0.001) and conduit strain (εe: 16.46%±3.96% and 11.30%±3.14% vs. 24.96%±5.27%; P<0.001). Booster pump strain (εa) was preserved in HTN (normal LAP) (17.45%±3.90% vs. 17.33%±3.13%; P>0.05) but reduced in HTN (elevated LAP) (13.47%±3.24%; P<0.001). Volumetric analysis revealed that maximum left atrial volume index (LAVImax) was comparable between controls and HTN (normal LAP) (P>0.05), while minimum left atrial volume index (LAVImin) showed a significant stepwise increase from controls (12.23±2.74 mL/m(2)) to HTN (normal LAP) (14.68±4.38 mL/m(2)) and HTN (elevated LAP) (21.28±8.74 mL/m(2); P<0.05). Multivariable analysis confirmed independent correlations of εs and εe with e' velocity (β=0.340/0.670), LAVImin (β=-0.231/-0.136), and global longitudinal strain (GLS) (β=0.468/0.380; all P<0.05); εa correlated inversely with E/e' (β=-0.192, P<0.05). ROC analysis identified LA strain parameters, particularly εs, as the strongest discriminator for HTN with elevated LAP [cutoff = 29.55%; the area under the curve (AUC) =0.896; sensitivity 84.4%, specificity 82.1%], with εs + LAVImin achieving superior diagnostic accuracy (AUC =0.919; sensitivity 89.6%, specificity 80.4%). CONCLUSIONS: CMR-derived LA strain demonstrates high discriminatory power for cardiac impairment in HTN patients, serving as a sensitive marker for LVDD with elevated LAP. LA strain and LAVmin show strong potential as clinically applicable parameters for routine diastolic function assessment in CMR.