Parameters to Improve Prediction of Atrial Fibrillation Development in HFpEF

Factors to improve the prediction of atrial fibrillation in patients with HFpEF.

Peak atrial longitudinal strain (PALS) and peak atrial contraction strain (PACS) were found to improve the prediction of atrial fibrillation (AF) occurrence in patients with heart failure with preserved ejection fraction (HFpEF), according to results from a prospective cohort study published in the Journal of the American College of Cardiologists: Cardiovascular Imaging.

Patients (n=170) with symptoms of HF admitted to a University Hospital in Poland between 2011 and 2014 were recruited for this study. A validation cohort of patients with HFpEF (n=103) were recruited from the Karolinska-Rennes study. Patients were assessed using exercise testing, echocardiography, blood analysis, and for the presence of AF through 2018.

Patients who developed (n=39) vs did not develop AF were significantly older (P =.01), had higher levels of brain natriuretic peptide (P =.001) and creatinine (P =.02), and higher Cohorts for Heart and Aging Research in Genomic Epidemiology AF (CHARGE-AF; P <.001) and CHA2DS2VASc (P <.001) scores.

Patients who developed AF differed for most echocardiograph characteristics compared with patient without AF, including left atrial (LA) volume index (39.9±8.1 vs 33.6±9.3 ml/m2, respectively; P <.001), E/e’ ratio (13.6±5.3 vs 11.0±2.8, respectively; P <.001), E/a ratio (1.21±0.78 vs 0.87±0.29, respectively; P <.001), PALS (23.1%±6.5% vs 29.0%±7.4%, respectively; P <.001), and PACS (10.9%±3.7% vs 14.5%±4.0%, respectively; P <.001).

The strongest predictors of AF development were: PACS (area under the curve [AUC], 0.76; P <.001), LA volume index (AUC, 0.72; P <.001), PALS (AUC, 0.71; P <.001), E/e’ ratio (AUC, 0.65; P =.003), CHA2DS2VASc (AUC, 0.65; P =.003), and CHARGE-AF (AUC, 0.65; P =.004).

In a prediction model in which CHA2DS2VASc, E/e’ ratio, and left atrial volume index were included, AF prediction improved by 37% (P =.04) with the addition of PALS and by 63% (P <.001) with the addition of PACS to the model.

Patients with PACS £12.7% were found to be at highest risk for AF, and those with PACS >12.7% and PALS £29.4% were at low risk for AF but high risk for left atrial volume index >34.3 mL/m2. These cut-offs had a sensitivity of 97% and specificity of 60% for predicting AF.

This study may have underestimated the rate of AF due to the relatively short follow-up period.

“The combination of atrial remodeling characteristics (LA deformation and LA size) may offer a sensitive tool to guide screening for AF risk in this population,” concluded the study authors.


Jasic-Szpak E, Marwick T H, Donal E, et al. Prediction of AF in Heart Failure With Preserved Ejection Fraction. Incremental Value of Left Atrial Strain. JACC Cardiovasc Imaging. 2021;14(1):131-144. doi:10.1016/j.jcmg.2020.07.040