Abstract Introduction In acute heart failure (AHF) patients, non-invasive estimation of left ventricular filling pressures (LVFPs) appears crucial to guide management. Although poorly investigated, left atrial (LA) mechanics play a pivotal role in this setting. Interestingly, recent findings indicate that combining LARS with conventional diastolic echocardiographic parameters, such as E/e’, may yield even more reliable estimations of LVFPs. Purpose This report sought to assess the correlation of echocardiographic LA stiffness index with invasive LVFPs and its diagnostic accuracy as compared to other parameters used in clinical practice. Methods In this observational, prospective study, 104 patients with suspected acute coronary syndrome and signs/symptoms of AHF were enrolled. Available invasive estimation of LVFPs was required. Comprehensive echocardiography was performed for all patients, including LA reservoir strain (LARS). LA stiffness index was derived as early diastolic transmitral inflow velocity/mitral annulus early diastolic velocity [E/e']/LARS. Predictive models based on LA stiffness index and LARS were built and tested in independent populations by randomly assigning patients to a "training group" for model building and a "test group" for model evaluation, using simple random sampling. Sampling was carried out with the "sample" function of R, with a 4:1 group ratio. Models’ performances were evaluated in the independent population "test group" against the guideline-recommended echocardiographic flowchart for the assessment of diastolic dysfunction. Results Invasively measured LVFPs showed a strong correlation with LA stiffness index (Spearman ρ = 0.773, p < 0.0001), as well as with LARS and E/e’. Receiver operating characteristic (ROC) curve analysis was used to demonstrate better accuracy performed by LA stiffness index than other echocardiographic metrics in predicting high LVFPs. Guideline-recommended assessment of diastolic function was finally compared to LARS and LA stiffness index performances in an independent population group; we were hence able to obtain a LA stiffness threshold of 0.48 with a positive predictive value of 91.7% and a negative predictive value of 88.9% in identifying patients with high LVFPs. Conclusions For the first time, diagnostic performance of LA stiffness index has been investigated in a heterogeneous AHF population, providing correlations with invasively measured LVFPs and comparisons with established diastolic function metrics.
Read full abstract