Abstract 14883: Pulmonary Vein Remodeling After PVI; Insights From Cardiac MRI
Background: Pulmonary vein remodeling has been proposed as a mechanism leading to sustaining of normal sinus rhythm following successful pulmonary vein isolation for atrial fibrillation (AF).
Hypothesis: Assuming a robust and reproducible technique for pulmonary vein (PV) measurement, we hypothesized that PV anatomy would finitely parallel and potentially predict PVI response rate.
Methods: We evaluated 91 patients with AF who had CMR before and 6±2 months after PVI. Each pulmonary vein cross sectional areas (CSA) and LA volume was measured from the 3D MRA of the left atrium and pulmonary veins within 1cm of the ostium using triangulation technology. Patients were divided into responders (R) and non responders (NR), depending on two 15-day Holter monitor and were classified as (R) if they had none or < 1 minute of AF during monitoring period.
Results: Using a 3D triangulation MRA technique, all (100%) of the 364 PV’s were successfully measured. Holter-defined 66 patients (73%) as R while 25 (27%) were NR with similar mean age; BSA, PVI technique and antiarrhythmic therapies (p=NS). While all four pulmonary veins in each patient and the LA volume were significantly reduced in size post procedure in both groups (15±30 vs. 9±30%, p<0.05) and (9±20 vs. 7±20%, p<0.05), respectively, no predictive power towards R or NR was found. Similarly, this reduction neither predicted the reduction in 3D LA volume, LVEF or the degree of MR. Further, predictive power was not augmented by normalization of PV by BSA, LVEDV and LA volume. Cluster analysis was performed and a scatter plot used to show PV relationship to recurrence, however, no particular linear pattern was found. (R2=0.25).
Conclusion: Despite the clearly favorable effects of PVI on LV and LA remodeling Cluster analysis demonstrates that AF recurrence is likely far more due to physiologic and electrical factors rather than technical (abalation approaches) issues. parameters, classically observed parallel remodeling effects on PV anatomy, despite multiple statistical approaches, was not present. Likewise, PV size remarkably failed to predict PVI response rate. This indicates that PV size counterintuitively plays a passive role in maintenance of NSR, potentially acting merely as an epiphenomenon.
Author Disclosures: S.S. Khidr: None. S.T. Reddy: None. M. Doyle: None. V. Farah: None. D. Vido: None. R. Lombardi: None. R. Williams: None. J. Yamrozik: None. W. Belden: None. M. Abdel Ghany: None. H.H. Elaraby: None. D. Fouad: None. M. Shah: None. R.W. Biederman: None.
- © 2016 by American Heart Association, Inc.