We congratulate Hu et al for reporting an important multicenter randomized trial evaluating intracardiac echocardiography versus transesophageal echocardiography for thrombus screening before atrial fibrillation ablation. Multiple key points merit further discussion.
First, the trial excluded patients with left ventricular ejection fraction <30%, New York Heart Association III to IV heart failure, and recent major cardiovascular events (e.g., acute myocardial infarction within 3 months). These groups are regularly encountered in real-world practice and have a greater left-atrial appendage thrombus prevalence. The mean left-atrial diameter was approximately 43 mm in both groups, which is consistent with mild enlargement. The current study population may not fully represent real-world practice.
Second, the observed periprocedural thromboembolic event rates were lower than those assumed for the trial’s sample-size estimation, which increases statistical fragility. This is further compounded by using a relatively wide noninferiority margin of 0.8%—which was not mentioned in the original trial registration—despite an expected absolute difference of 0.1% between the two groups. Additionally, the primary endpoint was driven largely by transient ischemic attack (TIA), a diagnosis with variable clinical interpretation and susceptibility to misclassification. The lower-than-expected event rate, the subjective nature of TIA diagnosis, and the wide noninferiority margin render the conclusions vulnerable to uncertainty.
Third, the trial generalizability is further limited by the anticoagulant profile and the absence of baseline renal function. Nearly all patients received rivaroxaban, whereas apixaban is the most used DOAC (Direct Oral Anticoagulant) in the United States and many European countries. , Because renal clearance influences DOAC pharmacokinetics, thromboembolic, and bleeding risk, the absence of baseline renal function limits extrapolation to real-world practice.
In summary, the trial provides valuable data supporting intracardiac echocardiography-guided screening in selected low-risk patients. Exclusion of high-risk populations, the lower-than-expected event rates, a wide noninferiority margin applied to subjective outcomes like TIA, and lack of baseline renal function warrant cautious interpretation of the results. Future trials must address these limitations to better inform our clinical practice.
CRediT authorship contribution statement
Bharat Rawlley: Conceptualization, Investigation, Methodology, Writing– original draft. Kartik Gupta: Writing– review & editing, Writing– original draft.
Declaration of competing interest
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
