Introduction: Differentiation between heart failure with preserved ejection fraction (HFpEF) and isolated diastolic dysfunction (DD) at rest is crucial, since the prognosis is different. Symptoms are often non-speci-fic, while NTproBNP might not be available. Increased NTproBNP is predicted by none of the currently used transthoracic echocardiographic (TTE) parameters. However, assessment of left atrial (LA) function by speckle tracking echocardiography (STE) might be a potential new marker of increased LV filling pressure.

Objective: To assess LA function by STE in HFpEF and DD, on top of the currently used TTE parameters, in order to establish the added value of LA deformation in the diagnosis of HFpEF.

Methods: 70 patients were enrolled prospectively: 40 with HFpEF (68 ± 9 yrs) and 30 with DD (60 ± 10 yrs). TTE was used to assess LV ejection fraction (LVEF), E/E’ ratio, left atrial volume index (LAVi), and systolic pulmonary arterial pressure (sPAP). STE was used to assess LA functions: reservoir function by strain from MVC to MVO (LASr) and positive strain rate (LASRr), conduit function by strain from MVO to onset of atri-al contraction (LAScd) and early negative strain rate during conduit phase (LASRcd), and LA pump functi-on by negative strain at MVC (LASct) and late negative strain rate during atrial contraction phase (LASRct). NTproBNP was measured in all patients.

Results: HFpEF patients had significantly higher LVEF (62 ± 6 vs. 57 ± 8, p<0.008), NTproBNP (329 ± 383 vs. 37 ± 26, p<0.001), E/E’ ratio (10.4 ± 2.7 vs. 7.5 ± 1.8, p<0.001), and sPAP (34 ± 11 vs. 23 ± 7, p<0.001), but similar LAVi (40 ± 9 vs. 39 ± 11) compared to DD, suggesting higher LV filling pressure. LA reservoir and conduit function were similar AS (LASr 25 ± 4 vs. 26 6, p=0.4, LASRr -1.28 ± 0.3 vs. -1.26 ± 0.3, p=0.8, LAScd 11.6 ± 5.5 vs. 11.2 ± 3.6, p=0.7, LARScd -1.37 ± 0.5 vs. -1.5 ± 0.6, p=0.4). However, LA pump function was significantly lower in HFpEF, expressed by LASRct (-1.07 ± 0.6 vs. -1.76 ± 0.7, p<0.001). NTproBNP cor-related with E/E’ ratio, sPAP, and LASRct (all r=0.44, p<0.001), but not with LAVi. By multiple regression analysis, best predictor for NTproBNP>125pg/ml was LASRct (r=0.60, r2=0.30, p<0.001). LASRct <-1.29 (AUC=0.82, sensitivity 75%, specificity 81%) was the only predictor of NTproBNP>125pg/ml.

Conclusions: LA pump function is the only predictor of NTproBNP>125pg/ml. This parameter should be in-corporated in the current protocols for the diagnosis of HFpEF. This work was supported by a grant of Ministery of Research and Innovation, CNCS-UEFISCDI, project number PN-III-P1-1-TE-2016-0669, within PNCDI III