Atrial fibrillation (AF) is one of the most common clinical arrhythmias, with a prevalence rate of 0.4–2%, and increased with advancing age [1]. AF is associated with a fivefold increased risk of stroke; moreover, AF-related stroke is more severe than non-AF-related stroke [2]. Thus, adequate anticoagulant therapy is essential for patients with AF. Over the past few years, non-vitamin-K antagonist oral anticoagulants (NOACs) have been increasingly widely used and have reduced stroke and systemic embolism compared with warfarin, with a lower bleeding risk [3].

Radiofrequency catheter ablation (RFCA) is a safe and effective alternative method to sustain sinus rhythm and improve symptoms for patients with AF. However, the risk of periprocedural thromboembolism remains an intractable issue, with an incidence rate of 1–5%. The routine application of heparin to maintain an activated clotting time (ACT) of 250–350 s during RFCA was recommended in the Chinese AF treatment guidelines [4]. The Heart Rhythm Society's Scientific Statement suggested 50–100 U/kg of heparin administration before transseptal puncture in patients on warfarin anticoagulants. Compared with patients on warfarin, the average time required to reach the target ACT was longer in patients treated with NOACs, who needed more intraprocedural heparin doses [5]. Nevertheless, few studies focusing on the initial heparin dosage in patients with NOACs have been reported. We compared the actual percentages of measurements at the target ACT with different initial heparin doses in patients taking NOACs and patients taking warfarin to determine the appropriate dosage of initial intraprocedural heparin.

This study demonstrated that for AF patients receiving NOACs, an initial heparin dosage of 120 U/kg or 130 U/kg could provide adequate intraprocedural anticoagulation with a favorable safety profile, regardless of whether the NOAC was dabigatran or rivaroxaban.

Compared with warfarin, NOACs are noninferior in the prevention of strokes or systemic embolism and have a favorable safety profile; in addition, it is easier for patients adhere to NOAC treatment than to warfarin treatment [7]. This superiority was more evident in patients with poor INR control (TTR < 66%) or in Asians [8, 9]. In our study, there was an interesting phenomenon in which the WG's body weight was lower than that in the NG, indicating that patients with high weight were more inclined to choose NOACs for adequate adherence [10]. Indeed, none of the patients enrolled were underweight, and the intraoperative heparin dose was based on body weight.

For AF patients undergoing RFCA, the increased risk of periprocedural thromboembolic and bleeding complications is an intractable issue. The occurrence of thromboembolism during RFCA may be due to multiple factors, including catheter manipulation, especially in the left atrium; endothelial injury; the formation of necrotic debris; hypercoagulability; the activation of coagulation triggered by intrinsic or extrinsic pathways; and altered blood flow after AF conversion [11]. Subsequent atrial stunts and decreased contractility followed by ablation procedures may be attributed to postprocedural thromboembolism [12]. Asymptomatic cerebral infarction may be a common thromboembolic event and is associated with the number of intraprocedural electrical cardioversions [13]. According to present studies and recent guidelines, it is preferable to administer therapeutic oral anticoagulants (OAC) for at least three weeks before ablation and two months after ablation and continue long-term use for patents with stroke risk (CHA₂DS₂-VASc score ≥ 2 in men or ≥ 3 in women) [14, 15]. Uninterrupted periprocedural anticoagulant treatment is superior to the interrupted strategy with intravenous heparin or subcutaneous low-molecular-weight heparin (LMWH) [16, 17]. Furthermore, uninterrupted NOAC treatment surpass warfarin treatment [18, 19]. Whether it is necessary to omit one or two NOAC doses before ablation is indeterminate. The regimens of the omission of one or two NOAC doses before ablation was not recommended in the 2020 ESC guidelines for AF because OAC treatment was truly uninterrupted in the related RCTs [14]. Gorla et al. [20] compared three strategies for periprocedural OAC administration: uninterrupted, mildly interrupted (< 12 h), and interrupted (> 12 h) and found no difference in the number of major bleeding and thromboembolic complications. However, the weighted mean incidence of overall bleeding was higher in the uninterrupted and mildly interrupted strategies than in the interrupted strategy, indicating that an interrupted approach may also be safe and effective.

A safe and adequate intraprocedural anticoagulation is essential during the ablation procedure, and a dose of heparin should be administered before or immediately following transseptal puncture. Then, heparin will be infused intravenously to achieve and maintain the target ACT [21]. Heparin's standard initial dosage was 50–100 U/kg, and 100 U/kg was more common for patients receiving warfarin [22]. For patients who are on NOACs, heparin's initial dosage is higher and these patients need more time to achieve the target ACT and more heparin to maintain their target ACT compared with patients anticoagulated with warfarin [23]. Yamaji et al. [24] found that adequate initial heparin dosages for AF ablation in patients on NOACs are 10–20% higher than those in patients on warfarin, and they indicated that 120 or 130 U/kg dabigatran, and 130 U/kg rivaroxaban and apixaban are adequate initial heparin dosages. These findings are consistent with ours. The ACTs in WG were higher and achieved the target ACT earlier than NGs. Moreover, more ACTs in WG were higher than 350 s, which was over the target ACT, resulting in a lower percentage of measurements at the target ACT. Kishima et al. [5] developed a new initial heparin administration regimen with heparin added every 30 min based on the measured ACT to raise the target ACT (≥ 300 s) achievement rate in patients on NOACs. For patients with baseline ACT > 130 s, an initial dose of “100U/kg + 3000U” was administrated, followed by intermittent intravenous heparin. For patients with a baseline ACT < 130 s, when the initial heparin dose was increased to “100 U/kg + 5000 U” from “100 U/kg + 3000 U”, the rate of achieving the target ACT rose to 80.5% from 41.6% with no increase in the number of bleeding complications.

Currently, there is no consensus on the optimal intraoperative target ACT value. As mentioned above, AF-related guidelines and recommendations advocate maintaining an ACT of 300–350 s; however, a lower target ACT (< 300 s) was evaluated in some studies [22, 25]. Some authors have shown that RFCA of AF could be performed under lower intraoperative target ACT values with no increased risk of stroke or thromboembolic events and reduced bleeding complications in their studies [26, 27]. According to our experiences and previous Chinese guideline [4], the target ACT in this study was set as 250–350 s. There was no thromboembolic complication detected in this study; thus, we think a conservative ACT target is reasonable based on more frequent ACT monitoring and the wide use of the saline-irrigated catheter. According to the recently widely accepted guideline, we compared the average percentage of ACTs in 300–350 s [21, 28]. No significant difference was detected; however, the percentage in the NG130 was higher than that in the WG, indicating the dosage of 130 U/kg can provide a more adequate intraprocedural anticoagulant effect.

Some researchers suggested that a continuous heparin intravenous infusion is superior to an intermittent infusion for maintaining a target ACT during the AF ablation [29]. The percentage of ACT measurements at the optimal ACT in continuous heparin administration was significantly higher than that in the intermittent strategy, with lower total heparin levels and more stable ACT levels. This might be due to the dose-dependent pharmacokinetics and short half-life of heparin [30]. In our study, the administration of intravenous heparin after the initial administration was discontinuous, and the dosage of additional heparin was based on the measured ACT value, which was monitored every 15 min. There was no increase in the incidence of bleeding or thrombotic complications. Furthermore, CA of AF is a well-established treatment, and the CA procedure can be completed within 120 min, especially when performed by well-trained operators. Thus, we consider that intermittent intraprocedural heparin infusion guided by more frequent ACT monitoring is feasible and safe.

The average total heparin dose used to manage ACT was slightly higher, and the mean ACT level was marginally lower in rivaroxaban than in warfarin in the VENTURE-AF trial [17]. This subtle difference in ACTs may vary with the selected NOAC. Kottmaier et al. [31] found that the baseline ACT and mean ACT were higher in patients who received uninterrupted edoxaban than in those who received warfarin. This reason is unclear and may be related to the small sample number in the edoxaban group and the conservative target ACT (280–300 s). We also found that the intraprocedural ACT levels were higher in the WG than in the NGs. However, no significant difference was detected between the WG and NG130, which may be due to a higher initial dose of heparin in NG130. The baseline ACT in the WG was higher than that in the NGs, but there was no significant difference in our study.

Bleeding complications of RFCA are infrequent; access-site hematoma, followed by pericardial effusion with tamponade, are common among them [32]. The incidence of hemorrhage in our study was approximately 2–5%, which is similar to that reported in previous research. The incidence of bleeding was slightly higher in the WG than in the NGs, especially for those who received a lower initial dosage of heparin. Although one patient occurred pericardiac effusion in NG130, the incidence was still low and might be due to the additional ablation (alation at mitral annulus). Notably, the ACTs in NG130 achieved the target earlier than in NG110 and NG120. Compared with warfarin, there was a trend favoring NOACs in terms of major bleeding complications during the ablation of AF [33].

This study is a single-center study with a small sample. There was no significant difference in the incidence of perioperative bleeding or thromboembolic complications among the four groups; therefore, we could not fully assess the safety and risk. The target ACT in this study is slightly lower than the latest guideline. Moreover, we only compared the anticoagulation effect of dabigatran at 110 mg in this study, and this does not indicate the effect of the dose of 150 mg dabigatran. Multicenter trials with a large sample are needed to verify this hypothesis further.

The initial heparin dosage and total heparin were higher for patients with AF on uninterrupted NOAC treatment who were undergoing RFCA than for patients taking warfarin. The appropriate initial heparin dosage was 120 U/kg or 130 U/kg for patients anticoagulated with NOACs. The dosage of 130 U/kg allowed ACT to reach the target earlier than the dosage of 120 U/kg.