Prediction of atrial fibrillation in patients with an implantable cardioverter-defibrillator and heart failure

• Published in: European journal of heart failure 2010-10, Vol.12 (10), p.1101-1110
• Main Authors: Bertini, Matteo, Borleffs, C. Jan Willem, Delgado, Victoria, Ng, Arnold C.T., Piers, Sebastiaan R.D., Shanks, Miriam, Antoni, M. Louisa, Biffi, Mauro, Boriani, Giuseppe, Schalij, Martin J., Bax, Jeroen J., Van de Veire, Nico R.L.
• Format: Article
• Language: English
• Subjects: Atrial fibrillation; Atrial Fibrillation - diagnostic imaging; Atrial Fibrillation - mortality; Atrial Fibrillation - pathology; Cohort Studies; Confidence Intervals; Defibrillators, Implantable; Disease Progression; Female; Heart Atria - diagnostic imaging; Heart failure; Heart Failure - diagnostic imaging; Heart Failure - mortality; Heart Failure - pathology; Humans; Italy; Kaplan-Meier Estimate; Male; Middle Aged; Multivariate Analysis; PA-TDI duration; Prognosis; Proportional Hazards Models; Prospective Studies; Regression Analysis; Risk Assessment; Time Factors; Ultrasonography, Doppler
• ISSN: 1388-9842; 1879-0844
• DOI: 10.1093/eurjhf/hfq126

Aims Heart failure and atrial fibrillation (AF) frequently coexist and AF worsens heart failure prognosis. Device‐based diagnostics derived from implantable cardioverter‐defibrillator (ICD) interrogation provide an accurate method for detecting AF episodes. This study sought to determine clinical and echocardiographic predictors of AF occurrence, including an index of total atrial conduction time derived by tissue Doppler imaging (PA‐TDI duration), in patients with heart failure. Moreover, the role of PA‐TDI duration on the prediction of AF occurrence in subgroups of patients with and without history of AF was explored. Methods and results A cohort of 495 heart failure patients who underwent ICD implantation was studied. Baseline echocardiographic parameters of systolic and diastolic function were evaluated together with clinical parameters. Furthermore, PA‐TDI duration was measured. All patients were prospectively followed up after ICD implantation for AF occurrence detected by ICD interrogation. A total of 142 (29%) patients experienced AF over a follow‐up period of 16.4 ± 11.2 months. PA‐TDI duration was longer in patients with AF occurrence when compared with patients without AF occurrence (154 ± 27 vs. 135 ± 24 ms, P < 0.001). On Cox‐multivariable analysis, female gender [hazard ratio = 1.60; 95% confidence intervals (CI) = 1.09–2.35; P = 0.017], history of AF (hazard ratio = 2.22; 95% CI, 1.51–3.27; P < 0.001), and PA‐TDI duration (hazard ratio = 1.27; 95% CI, 1.13–1.42; P < 0.001) were independent predictors of AF occurrence. In the subgroups of patients with and without history of AF, PA‐TDI duration remained an independent predictor of AF occurrence. Conclusion PA‐TDI duration may be useful to risk‐stratify for AF occurrence in heart failure patients with and without a history of AF.