Highlight
– In a multicenter randomized trial of 1,810 patients undergoing catheter ablation for atrial fibrillation, ICE was noninferior to TEE for preventing periprocedural thromboembolic events (0.4% vs 0.6%).
– ICE was associated with fewer major bleeding events related to transseptal puncture, substantially reduced fluoroscopy time and preprocedural waiting time, and improved patient-reported anxiety/depression.
Background
Catheter ablation is an established rhythm-control strategy for symptomatic atrial fibrillation (AF). Before left atrial instrumentation, exclusion of preexisting intracardiac thrombus—particularly in the left atrial appendage (LAA)—is a standard safety step to minimize the risk of procedural stroke or systemic embolism. Transesophageal echocardiography (TEE) has long been considered the reference standard for thrombus detection prior to AF ablation, but it requires esophageal probe placement and usually procedural sedation or anesthesia, and carries small but important risks (esophageal trauma, aspiration, and patient discomfort).
Intracardiac echocardiography (ICE) offers real-time imaging from within the heart, enabling visualization of the LAA, interatrial septum, and catheters during transseptal puncture and ablation. Observational studies and single-center series have suggested that ICE can reduce fluoroscopy exposure and may improve procedural safety, but high-quality randomized evidence comparing ICE and TEE for preablation thrombus screening has been limited.
Study design
The ICE vs TEE Study Investigators reported a pragmatic multicenter randomized clinical trial conducted at 10 hospitals in China between August 2022 and July 2023 (ClinicalTrials.gov Identifier: NCT05466266). Adults scheduled for catheter ablation of AF who met prespecified eligibility criteria were randomized 1:1 to thrombus screening with ICE (n = 906) or TEE (n = 904) prior to ablation. The primary endpoint was the incidence of periprocedural thromboembolic events (stroke, transient ischemic attack, or systemic embolism) assessed through 30 days after the procedure. Secondary endpoints included detection of intracardiac thrombus, procedural safety (including major bleeding related to transseptal puncture), procedural efficiency metrics (fluoroscopy time, preprocedural waiting time), and patient-reported measures such as anxiety or depression prevalence. Data analysis occurred from August to December 2023.
Key findings
Baseline characteristics were balanced between groups (mean age 64.3 years, 48.0% women, 49.0% paroxysmal AF). Primary and secondary outcome results were as follows.
Primary outcome: thromboembolic events
Periprocedural thromboembolic events occurred in 4 of 906 patients in the ICE group (0.4%) and 5 of 904 patients in the TEE group (0.6%). The risk difference was -0.11% with a Farrington-Manning 95% confidence interval of -0.84% to 0.62%; P for noninferiority = .01. These low, numerically similar event rates support the conclusion that ICE is noninferior to TEE for prevention of short-term thromboembolic complications in this context.
Thrombus detection
Intracardiac thrombus was detected in 2.0% of patients assessed with ICE and 1.5% with TEE (relative risk [RR] 1.29; 95% CI, 0.64–2.61; P = .48). Although overall detection rates were similar, ICE identified a higher proportion of non–left atrial appendage thrombi (0.6% vs 0%; P < .001). This finding suggests ICE may improve visualization of certain intracardiac regions not easily imaged with TEE, but the clinical implications (for example, whether these represent true thrombi of embolic relevance) require further evaluation.
Procedural safety
Major bleeding related to transseptal puncture was less frequent with ICE (0.2% vs 1.2%; RR 0.18; 95% CI, 0.04–0.81; P = .03). The lower bleeding rate plausibly reflects direct intraprocedural visualization of the septum and adjacent structures with ICE, enabling safer transseptal access.
Procedural efficiency and patient experience
ICE substantially reduced mean fluoroscopy time (4.2 ± 1.5 vs 9.3 ± 3.0 minutes; P < .001) and preprocedural waiting time (14.4 ± 8.0 vs 23.6 ± 10.5 hours; P < .001). The prevalence of anxiety or depression was lower in the ICE group (24.6% vs 37.5%; RR 0.66; 95% CI, 0.56–0.76; P < .001), reflecting better patient comfort and possibly less need for anxiolytic or general anesthesia interventions.
Clinical interpretation
This large randomized trial provides the most rigorous evidence to date that ICE can serve as a safe and effective alternative to TEE for preablation thrombus screening in patients undergoing AF ablation. The primary endpoint event rates were low in both arms, and noninferiority of ICE was met with a narrow confidence interval around the small absolute difference. Beyond equivalence for embolic prevention, ICE offered pragmatic advantages relevant to procedural workflow and patient-centered care: fewer serious transseptal bleeding events, substantially less fluoroscopy exposure, shorter preprocedural wait times, and improved patient comfort.
In practice, reduced fluoroscopy time is attractive given cumulative radiation concerns for patients and staff. Shorter preprocedural waiting time may reflect the ability to perform ICE at the time of the ablation without arranging separate preprocedural TEE under sedation or anesthesia and could improve throughput. Lower anxiety/depression prevalence likely reflects avoidance of an invasive esophageal probe and reduced need for conscious sedation or general anesthesia.
Expert commentary and guideline context
Current major guidelines recognize the importance of excluding LAA thrombus prior to LA instrumentation but do not mandate a specific imaging modality. The 2020 European Society of Cardiology (ESC) guidelines on atrial fibrillation and the 2017 HRS/EHRA/ECAS expert consensus statement both discuss imaging strategies for thrombus detection and highlight TEE as the traditional standard. However, these documents also acknowledge alternative imaging approaches where appropriate expertise and equipment exist.
The trial’s findings align with the growing procedural adoption of ICE in electrophysiology laboratories, moving beyond its role for catheter guidance to a more central role in preprocedural safety evaluation. ICE may be particularly useful where TEE is logistically challenging, when general anesthesia is contraindicated, or where minimizing staff exposure to radiation and shortening procedural pathways is a priority.
Limitations and generalizability
Important limits should frame interpretation. The trial enrolled patients at 10 centers in China; patterns of sedation, anesthesia, and resource allocation vary internationally, and cost structures for disposable ICE catheters differ across health systems. Operator experience with ICE was not detailed in the summary; outcomes may be optimized in centers with established ICE expertise but different in centers new to the technology.
The absolute incidence of periprocedural thromboembolism was low in both groups, which is reassuring but limits power to detect very small absolute differences. The summary did not specify the prespecified noninferiority margin; readers should consult the full report for details of the statistical plan. Blinding of operators to imaging modality is inherently impossible and could influence some procedural decisions (detection bias), although the primary endpoint—clinical thromboembolism—was objective.
The increased detection of non-LAA thrombi with ICE is intriguing but requires cautious interpretation: whether these findings represent true thrombus with similar embolic risk as LAA thrombus, or represent imaging artifacts or clinically irrelevant findings, is uncertain. Finally, cost-effectiveness was not evaluated; ICE catheters are single-use devices with significant per-case cost, though these may be offset by reductions in anesthesia use, TEE-related resources, or procedure cancellations.
Implications for practice and future research
For electrophysiology teams with access to ICE and clinicians skilled in its use, this trial supports adopting ICE as a viable primary strategy for preablation thrombus screening. Health systems should weigh local cost, operator expertise, and logistics. Training programs should incorporate ICE imaging interpretation for operators performing AF ablation.
Open questions to guide future work include: formal cost-effectiveness analyses comparing ICE and TEE workflows in diverse health systems; longer-term follow-up beyond 30 days for embolic or bleeding outcomes; subgroup analyses by AF type, anticoagulation status, and LAA morphology; and studies assessing the clinical relevance of non-LAA thrombi identified by ICE.
Conclusion
This randomized multicenter trial demonstrates that ICE is noninferior to TEE for preventing periprocedural thromboembolic events in AF ablation, while offering additional advantages in procedural safety (fewer transseptal bleeding events), efficiency (reduced fluoroscopy and shorter waiting time), and patient comfort. Adoption of ICE may be appropriate where institutional resources and expertise allow, though local cost-effectiveness and operator training remain important considerations.
Funding and clinicaltrials.gov
Trial registration: ClinicalTrials.gov Identifier: NCT05466266. Funding sources and detailed sponsor declarations are available in the full publication (Hu et al., JAMA Cardiol. 2025).
References
1. Hu X, Jiang W, Wang X, et al; ICE vs TEE Study Investigators. Intracardiac vs Transesophageal Echocardiography in Atrial Fibrillation Ablation: A Randomized Clinical Trial. JAMA Cardiol. 2025 Oct 8:e253687. doi: 10.1001/jamacardio.2025.3687 . PMID: 41060665 ; PMCID: PMC12509080 .
2. Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation. Eur Heart J. 2020; (guideline document).
3. Calkins H, Hindricks G, Cappato R, et al. 2017 HRS/EHRA/ECAS/APHRS/SOLAECE Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation. Heart Rhythm. 2017;14(10):e275–e444.