Stef Zeemering1, Theo A R Lankveld1,2, Pietro Bonizzi3, Ione Limantoro2, Sebastiaan C A M Bekkers2, Harry J G M Crijns2, Ulrich Schotten1. 1. Department of Physiology, Maastricht University, Universiteitssingel 50, ER Maastricht, The Netherlands. 2. Department of Cardiology, Maastricht University Medical Center, AZ Maastricht, The Netherlands. 3. Department of Data Science and Knowledge Engineering, Maastricht University, Bouillonstraat 8-10, LH Maastricht, The Netherlands.
Abstract
Aims: Non-invasive characterization of atrial fibrillation (AF) substrate complexity based on the electrocardiogram (ECG) may improve outcome prediction in patients receiving rhythm control therapies for AF. Multiple parameters to assess AF complexity and predict treatment outcome have been suggested. A comparative study of the predictive performance of complexity parameters on response to therapy and progression of AF in a large patient population is needed to standardize non-invasive analysis of AF. Methods and results: A large variety of ECG complexity parameters were systematically compared in patients with recent onset AF undergoing pharmacological cardioversion (PCV) with flecainide. Parameters were computed on 10-s 12-lead ECGs of 221 patients before drug administration. The ability of ECG parameters to predict successful PCV and progression to persistent AF (mean follow-up 49 months) was evaluated and compared with common clinical predictors. Optimal prediction performance of successful PCV using only one ECG parameter was low, using dominant atrial frequency [lead II, receiver operating area under curve (AUC) 0.66, 95% confidence interval [0.64-0.67]], but the optimal combination of several ECG parameters strongly improved predictive performance (AUC 0.78 [0.76-0.79]). While predictive value of the optimal combination of clinical predictors was low (AUC 0.68 [0.66-0.70], using right atrial volume and weight), adding ECG parameters strongly increased performance (AUC 0.81 [0.79-0.82], P < 0.001). Interestingly, higher dominant frequency and higher f-wave amplitude were associated with increased risk of progression to persistent AF during follow-up. Conclusion: Assessment of AF complexity from 12-lead ECGs significantly improves prediction of successful PCV and progression to persistent AF compared with common clinical and echocardiographic predictors.
Aims: Non-invasive characterization of atrial fibrillation (AF) substrate complexity based on the electrocardiogram (ECG) may improve outcome prediction in patients receiving rhythm control therapies for AF. Multiple parameters to assess AF complexity and predict treatment outcome have been suggested. A comparative study of the predictive performance of complexity parameters on response to therapy and progression of AF in a large patient population is needed to standardize non-invasive analysis of AF. Methods and results: A large variety of ECG complexity parameters were systematically compared in patients with recent onset AF undergoing pharmacological cardioversion (PCV) with flecainide. Parameters were computed on 10-s 12-lead ECGs of 221 patients before drug administration. The ability of ECG parameters to predict successful PCV and progression to persistent AF (mean follow-up 49 months) was evaluated and compared with common clinical predictors. Optimal prediction performance of successful PCV using only one ECG parameter was low, using dominant atrial frequency [lead II, receiver operating area under curve (AUC) 0.66, 95% confidence interval [0.64-0.67]], but the optimal combination of several ECG parameters strongly improved predictive performance (AUC 0.78 [0.76-0.79]). While predictive value of the optimal combination of clinical predictors was low (AUC 0.68 [0.66-0.70], using right atrial volume and weight), adding ECG parameters strongly increased performance (AUC 0.81 [0.79-0.82], P < 0.001). Interestingly, higher dominant frequency and higher f-wave amplitude were associated with increased risk of progression to persistent AF during follow-up. Conclusion: Assessment of AF complexity from 12-lead ECGs significantly improves prediction of successful PCV and progression to persistent AF compared with common clinical and echocardiographic predictors.
Authors: Matthias Daniel Zink; Rita Laureanti; Ben J M Hermans; Laurent Pison; Sander Verheule; Suzanne Philippens; Nikki Pluymaekers; Mindy Vroomen; Astrid Hermans; Arne van Hunnik; Harry J G M Crijns; Kevin Vernooy; Dominik Linz; Luca Mainardi; Angelo Auricchio; Stef Zeemering; Ulrich Schotten Journal: Front Physiol Date: 2022-03-04 Impact factor: 4.566
Authors: Albert Dasí; Aditi Roy; Rafael Sachetto; Julia Camps; Alfonso Bueno-Orovio; Blanca Rodriguez Journal: Front Physiol Date: 2022-09-15 Impact factor: 4.755