The prognostic value of left atrial size and strain in predicting atrial fibrillation recurrence at 6 and 12 months after pulmonary vein isolation

Abstract Introduction Left atrial (LA) volume and LA strain (LAS) offer insights into structural and functional changes of the LA and may help predict atrial fibrillation (AF) recurrence after pulmonary vein isolation (PVI). Purpose To assess the predictive value of pre-procedural LA morpho-functional echocardiographic parameters in determining AF recurrence at 6 and 12 months post-PVI. Methods This single-center retrospective study analyzed AF patients who underwent initial PVI between 2015 and 2024, with pre-procedural echocardiographic assessment. LAS and volume measurements were reviewed, and AF recurrence at 6 and 12 months was determined using available 12-lead ECG and Holter monitoring. Receiver operating characteristic analysis assessed the predictive value of LA echocardiographic parameters. Results Of 216 patients undergoing PVI, 72 met inclusion criteria. Follow-up was completed by 60 at 6 months and 51 at 12 months. Median echocardiography-to-PVI interval was 312 days [IQR 120–610]. AF recurrence occurred in 9 (15.0%) at 6 months and 15 (29.4%) at 12 months. At 6 months, indexed LA diameter demonstrated the highest discriminatory power, with an AUC of 0.68 (p=0.21) and an optimal cutoff of >24.3 mm/m² (sensitivity: 57.1%; specificity: 91.7%). Indexed LA end-diastolic volume had an AUC of 0.64 (p=0.26) with an optimal cutoff of >41 mL/m² (sensitivity: 55.6%; specificity: 85%). Among LA strain parameters, LAS of conduit (LAScd) had the highest AUC (0.64, p=0.28) with a cutoff of ≤7.8% (sensitivity: 60%; specificity: 80.6%). LAS of contraction (LASct) and LAS of reservoir (LASr) demonstrated lower predictive performance, with AUCs of 0.54 (p=0.84) and 0.59 (p=0.59), respectively. At 12 months, indexed LA diameter showed a significant predictive value for AF recurrence, with an AUC of 0.72 (p=0.02), a cutoff of >23.6 mm/m², sensitivity of 53.9%, and specificity of 90.6%. Indexed LA end-diastolic volume had an AUC of 0.67 (p=0.07) with a cutoff of >33.5 mL/m² (sensitivity: 73.3%; specificity: 61.3%). Among strain parameters, LAScd had the highest AUC (0.66, p=0.11) with a cutoff of ≤14.3% (sensitivity: 100%; specificity: 37.5%). LASr demonstrated an AUC of 0.68 (p=0.13), a cutoff of ≤15.3% (sensitivity: 44.4%; specificity: 91.7%). LASct had the lowest predictive value, with an AUC of 0.57 (p=0.57). DISCUSSION: Despite some echocardiographic parameters demonstrating moderate discriminatory power, none reached robust statistical significance, except for indexed LA diameter at 12 months. The study's limitations likely contributed to these findings. A low inclusion rate introduced potential selection bias, and the small sample size may have limited the statistical power to detect significant associations. Additionally, the variable time between echocardiographic assessment and PVI could have affected the accuracy of pre-procedural measurements in predicting outcomes. Dropout rates further limit the ability to draw definitive conclusions.