Prediction of Left Atrial Fibrosis and Success of Catheter Ablation by Speckle Tracking Echocardiography in Patients Imaged in Persistent Atrial Fibrillation

Marchandise, Sébastien [UCL] Garnir, Quentin [UCL] Scavée, Christophe [UCL] Varnavas, Varnavas [UCL] Le Polain De Waroux, Jean-Benoît [UCL] Wauters, Aurélien [UCL] Beauloye, Christophe [UCL] Roelants, Véronique [UCL] Gerber, Bernhard [UCL]

(eng) BACKGROUND: Non-invasive evaluation of left atrial structural and functional remodeling should be considered in all patients with persistent atrial fibrillation (AF) to optimal management. Speckle tracking echocardiography (STE) has been shown to predict AF recurrence after catheter ablation; however in most studies, patients had paroxysmal AF, and STE was performed while patients were in sinus rhythm. AIM: The aim of this study was to evaluate the ability of STE parameters acquired during persistent AF to assess atrial fibrosis measured by low voltage area, and to predict maintenance of sinus rhythm of catheter ablation. METHODS: A total of 94 patients (69 men, 65 ± 9 years) with persistent AF prospectively underwent measurement of Global Peak Atrial Longitudinal Strain (GPALS), indexed LA Volume (LAVI), E/e' ratio, and LA stiffness index (the ratio of E/e' to GPALS) by STE prior to catheter ablation, while in AF. Low-voltage area (LVA) was assessed by electro-anatomical mapping and categorized into absent, moderate (>0 to <15%), and high (≥15%) atrial extent. AF recurrence was evaluated after 3 months of blanking. RESULTS: Multivariable regression showed that LAVI, GPALS, and LA stiffness independently predicted LVA extent after correcting for age, glomerular filtration rate, and CHA2DS2-VASc score. Of all the parameters, LA stiffness index had the highest diagnostic accuracy (AUC 0.85), allowing using a cut-off value ≥0.7 to predict moderate or high LVA with 88% sensitivity and 47% specificity, respectively. In multivariable Cox analysis, both GPALS and LA stiffness were able to significantly improve the c statistic to predict AF recurrence (n = 40 over 9 months FU) over CHARGE-AF (p < 0.001 for GPALS and p = 0.01 for LA stiffness) or CHA2DS2-VASc score (p < 0.001 for GPALS and p = 0.02 for LA stiffness). GPALS and LA stiffness also improved the net reclassification index (NRI) over the CHARGE-AF index (NRI 0.67, 95% CI [0.33-1.13] for GPALS and NRI 0.73, 95% CI [0.12-0.91] for LA stiffness, respectively), and over the CHA2DS2-VASc score (NRI 0.43, 95% CI [-0.14 to 0.69] for GPALS and NRI 0.52, 95% CI [0.10-0.84], respectively) for LA stiffness to predict AF recurrence at 9 months. CONCLUSION: STE parameters acquired during AF allow prediction of LVA extent and AF recurrence in patients with persistent AF undergoing catheter ablation. Therefore, STE could be a valuable approach to select candidates for catheter ablation.