Discussion
In this large cohort study which ran for over 10 years, the detection of LA thrombus at repeat sessions after initial catheter ablation for AF was very low (0.8%) in patients who underwent TEE examination and 0.5% in the total sample of patients who underwent repeat ablation. The use of imaging modalities at repeat ablation was increased when performed >180 days after the first procedure, and all patients who developed LA thrombi were scheduled to undergo repeat ablation >180 days after the initial ablation. No patients who did not undergo any imaging evaluation during the repeat session developed adverse events. Repeat ablation >180 days, non-paroxysmal atrial arrhythmia, and lower LVEF were independent predictors of thrombi development and severe SEC at repeat ablation in the multivariate analysis.
LA thrombi are unlikely to develop after initial catheter ablation for AF rather than prior to the first session, given the significantly decreased AF burden and continuous anticoagulant use after ablation.10-12 However, little is known about the prevalence of thrombus development and the requirements for imaging evaluation at repeat ablation are unclear. There is only one previous study demonstrating a 0.8% rate of thrombi detection in 263 patients who underwent repeat TEE within 365 days.13Nonetheless, all patients were prescribed warfarin, a vitamin K antagonist, in the previous study which was conducted over a long period of time. In contrast, an alternative anticoagulant, DOAC, which is mostly used in AF ablation, has the advantage of quickly promoting and maintaining its anticoagulant effect, which could support the suppression of thrombus development after ablation. Our study is unique in that it evaluated outcomes of using various imaging modalities in patients who received DOACs and warfarin, reflecting a recent trend in clinical practice.
However, despite the presumed safety and efficacy profiles of DOAC, three patients still developed LA thrombi at repeat ablation in our study, which indicates that a reduced risk remains and caution is still warranted at repeat ablation. More importantly, the thrombi developed even in patients with grade 0 SEC in the first session, and there is no guarantee of the absence of thrombi in reference to the examination findings obtained during the first session. Nonetheless, all these patients had a possible reason for the development of thrombi in our study, such as the long period until repeat ablation, inappropriate DOAC use, and decreased cardiac function prior to repeat ablation. In this regard, all patients who developed thrombi were scheduled to undergo repeat ablation >180 days after the first procedure, and the frequency of imaging evaluation was significantly increased when the repeat ablation was performed >180 days after the first procedure as compared to when performed at ≤ 180 days, which suggests that this cutoff duration value should be one of the safety parameters when considering the need for imaging evaluation to detect thrombi before repeat ablation. Moreover, a possible association with the risk of thrombus development in patients with decreased LVEF and non-paroxysmal atrial arrhythmia was mentioned in a previous report.13 Although a sub-therapeutic dose of warfarin should be avoided considering its significant relation to thrombus development despite less frequent use of warfarin to date, inappropriate DOAC dose use also increases the periprocedural risk of thrombi.13, 14 Nonetheless, with an expandable use of DOAC uninterruptedly during ablation, it may be possible to selectively screen patients to identify those at low risk of developing thrombi and those at high risk and who require TEE at repeat ablation.15 As long as patients with a low-risk profile continue on oral anticoagulants and undergo early repeat ablation, any imaging modality to determine the presence of LA thrombi may be omitted. Our results could prove useful in deciding whether imaging should be scheduled before repeat ablation according to each patient’s clinical profile.
In addition to TEE, contrast-enhanced CT and ICE are useful alternatives to exclude the possibility of LA thrombi, with the hope of obviating the need for TEE before ablation. In a meta-analysis of studies using delayed CT imaging protocols to screen for LA thrombi, the diagnostic accuracy for detection of LAA thrombi was reported to be nearly 100%.9 However, obtaining the CT image some time before the planned ablation day may limit its value, as indicated for Patient 2 in Table 4. On the other hand, previous data suggested that ICE could be valuable in safely and effectively detecting LA thrombi, whereas the result reflected in large part the variability seen when performing and interpreting the ICE imaging.16-18Moreover, all patients in which thrombi were detected in the present study had D-dimer levels within normal limits before repeat ablation;19 therefore, any imaging modality should still be used to screen for thrombi in patients at presumed risk of developing thrombi and those who cannot undergo TEE imaging.