Coagulation and heparin requirements during ablation in patients under oral anticoagulant drugs

Abstract Background Anticoagulation during catheter ablation should be closely monitored with activated clotting time (ACT). However vitamin K antagonists (VKA) or direct oral anticoagulant drugs (DOAC) may act differently on ACT and on heparin needs. The aim of this study was to compare ACT and heparin requirements during catheter ablation under various oral anticoagulant drugs and in controls. Methods Sixty consecutive patients referred for ablation were retrospectively included: group I (n = 15, VKA), group 2 (n = 15, uninterrupted rivaroxaban), group 3 (n = 15, uninterrupted apixaban), and group 4 (n = 15, controls). Heparin requirements and ACT were compared throughout the procedure. Results Heparin requirements during the procedure were significantly lower in patients under VKA compared to DOAC, but similar between DOAC patients and controls. Activated clotting time values were significantly higher in patients under VKA compared to DOAC and similar in DOAC patients versus controls. Furthermore, anticoagulation control as evaluated by the number/proportion of ACT> 300 as well as the time passed over 300 seconds was significantly better in patients under VKA versus DOAC, without significant differences between DOAC and controls. Finally, the number of patients/ACT with excessive ACT values was significantly higher in VKA versus DOAC patients versus controls. There was no significant difference between rivaroxaban and apixaban for ACT or heparin dosing throughout the procedure. Conclusion Vitamin K antagonists allowed less heparin requirement despite reaching higher ACT values and more efficient anticoagulation control (with more excessive values) compared to patients under DOAC therapy and to controls. There was no difference in heparin requirements or ACT between DOAC patients and controls.


| INTRODUC TI ON
Percutaneous catheter ablation has become the therapeutic of choice for patients with cardiac arrhythmias and the number of worldwide ablation procedures has dramatically increased over the previous years. 1 Anticoagulation is usually required during percutaneous catheter ablation procedures, because of the risk of material and lesion-induced cardiac thrombus and embolism. 2,3 This thrombo-embolic risk persists even in patients under oral anticoagulation, thus additional anticoagulation with unfractionated heparin is needed. 4 Because of combined risks of bleeding and thrombo-embolic events, it is furthermore recommended that anticoagulation levels should be closely monitored during ablation procedures: achieving and maintaining activated clotting time (ACT) > 300 seconds with heparin is currently considered mandatory for ablation of atrial fibrillation 5 and also commmonly used for ablation of ventricular arrhythmias. 6 Activated clotting time is a widely accepted coagulation monitoring technique 3,4,7 being perfectly adapted to clinical requirements, since results are very quickly available using bedside coagulation monitoring, allowing immediate adjustment of heparin needs. 8 In patients under long-term oral anticoagulation, ablations are optimally conducted without any interruption of anticoagulant drugs. 3 Uninterrupted use of warfarin has been previously recommended. 6 Feasibility and safety of uninterrupted direct acting oral anticoagulant (DOAC) on periprocedural periods of ablation has more recently been demonstrated [9][10][11] and is currently also recommended in the most recent guidelines. 6 Vitamin K antagonists (VKA) or DOAC may act differently on ACT, and thus on heparin needs for achieving ACT target values.
ACT and response to heparin have been shown to be different between patients receiving VKA or DOAC. [12][13][14] Furthermore, target ACT value in patients on uninterrupted DOAC is based on very limited evidence, mostly extrapolated from VKA-treated populations.
Nonetheless, differences in ACT and in heparin requirements according to the type of oral anticoagulant drug remain poorly investigated. Moreover no comparison with controls (ie, without previous oral anticoagulation) is currently available.
The aim of this study was to investigate and compare the ACT values and the requirement of heparin in patients undergoing percutaneous catheter ablation under various oral anticoagulant drug regimens and in controls.

| ME THODS
Sixty consecutive patients referred for ablation of left atrial or ventricular arrhythmias over the last six months at our center were retrospectively included. Patients were divided in four groups: group I (n = 15, patients under long-term VKA therapy), group 2 (n = 15, long-term rivaroxaban), group 3 (n = 15, long-term apixaban), and group 4 (n = 15, controls, no previous oral or intravenous/subcutaneous anticoagulation). The number of included patients in each group was similar for statistical purposes, and inclusions were consecutive, not random. Every patient from group 1 to 3 were on long-term oral anticoagulation therapy (for >2 months), with INR in the therapeutic range (group 1) or without any missing drug intake during the weeks before ablation (groups 2 and 3).
Patients underwent radio-frequency or cryo-ablation for left atrial arrhythmias (n = 45, groups 1 to 3) or ventricular arrhythmias (n = 15, controls). In every patient undergoing left atrial ablation, transoesophageal echocardiography ruled out atrial thrombus before the procedure. Values of each ACT and heparin dosing could be retrieved from our electrophysiological laboratory files since all these data were kept together with precise timings.
Automated ACT measurements were done using Hemochron ® Jr Signature + device (Accriva Diagnostics, San Diego, CA). No baseline ACT had been performed, but baseline prothrombin time and activated partial thromboplastin time (aPTT) were available in all patients. Hemostasis assays were performed using Neoplastine CI+ (prothrombin time) and CK Prest on Star Evolution (aPTT) both from STAGO (France).
Intravenous unfractionated heparin was infused as soon as the transseptal access was achieved (for patients with left atrial ablation or when transseptal access was needed for left ventricular ablation) or after arterial vascular access (in case of retrograde aortic access for left ventricular ablation). Initial heparin dosing was determined according to the current oral anticoagulation and/ or weight (usually 5000-7000 IU for patients under VKA and between 5000 and 10 000 IU for DOAC patients, based on weight and on our previous experience, or 100 IU/kg for non anticoagulated patients). First ACT was checked 20 minutes after the initial heparin infusion, then rechecked again 20 minutes later, once again 20 minutes later, then every 30 minutes. Target ACT value was 300 seconds in all patients. 3,4 Any ACT < 300 seconds was leading to new heparin infusion (dosing according to the ACT value, from 1000 to 3000 U).
The total amount of infused heparin during the whole procedure was calculated, as well as the time and heparin dosing needed before achieving ACT value of 300 seconds, the time passed over 300 seconds, the number/proportion of ACT ≥ 300 seconds, number/proportion of ACT and patients with ACT ≥ 400 seconds, number/proportion of ACT or patients with ACT < 300 after being ≥ 300 seconds.
Protamine was infused at the end of the procedure prior to removal of vascular sheaths in each patient (dosing according to the total heparin delivered). Hemorragic or embolic complications were

| S TATI S TI C S
Continuous variables are reported as median and IQR and compared with nonparametric tests (Mann Whitney when comparing two groups and Kruskal-Wallis when comparisons included more than two groups). Categorical variables were compared using Fischer's exact test. Comparisons of repeated ACT and of repeated heparin dosings between goups were performed using repeated measure ANOVA when there was sufficient number of values. Multivariate analysis was performed by logistic regression using variables significantly associated in univariate analysis, leading to calculation of odds ratio (OR) and 95% confidence intervals (CI).
A P < .05 was considered statistically significant for each analysis.

| Clinical characteristics
Clinical characterictics of the patient population are depicted in Table 1. Drugs used in group 1 were warfarin (n = 3), fluindione (n = 11) and acecoumarol (n = 1). All patients had INR in the therapeutic range before ablation. There was no missing drug intake during the weeks before ablation in groups 2 and 3 patients, and DOACs were "truly" uninterrupted before the procedure (not any discontinuation in any patient even the day of the procedure). Drug dosing in groups 2 and 3 were considered optimal according to age, weight and renal function: 20 mg rivaroxaban daily in each and 5 mg bid apixaban in each but one 82-year-old patient weighing 56 kg (2.5 mg bid). Data from this patient were not different from other group 3 patients, thus they were retained for analysis.
Comparisons between group 2 and 3 failed to demonstrate any significant difference, except for a trend towards higher CHA2DS2VaSC score in the rivaroxaban group (median 3, IQR 2.7) compared to that in apixaban group (median 1, IQR 1.7) (P = .09) ( Table 2). Thus, group 2 and 3 were mixed (DOAC group) and compared to group 1 (VKA) and controls (Table 3).
Controls were heavier compared to VKA and DOAC. There were more women in the VKA group compared to DOAC and controls. CHA2DS2VaSC score was higher in the VKA group as a result of more stroke and women in this group and bordeline older age.
Finally, glomerular filtration was lower in VKA patients compared to DOAC and controls.

| Coagulation and heparin requirements
Except for longer baseline aPTT in patients with rivaroxaban (median 1.2, IQR 0.1) compared to apixaban patients (median 1.1, IQR 0.1) (P = .005), there was no difference between group 2 and 3 patients for any coagulation time or heparin dosing at baseline or throughout the procedure (Table 4). Thus, group 2 and 3 were mixed (DOAC group) and compared to group 1 (VKA) and controls. Coagulation parameters and heparin requirements in patients under VKA, DOAC and controls are depicted in Table 5.
Baseline prothrombin time and aPTT were significantly lower in controls compared to DOAC, which were also significantly lower compared to VKA patients. Duration of the procedure was longer in controls but did not differ between VKA and DOAC patients.
Initial heparin dosing, dosing till achievement of ACT > 300 and total heparin requirements were significantly higher in controls   In previous large series however, no difference in embolic or bleeding complications had been noted between patients undergoing ablation procedures while on VKA or on uninterrupted DOAC therapy. [10][11][12]15,16 In meta-analysis, there was no significant association between bleeding events and ACT levels in patients receiving DOAC, while significant associations were found for embolic events. DOAC used in this study were both oral direct FXa inhibitors. 18 At first glance there is no reason for ACT to not be able to monitor DOAC effect while being a reliable marker of heparin activity, because inhibition of factor Xa is the main effect of these drugs. However ACT has been showed to be poorly correlated with DOAC therapy, with large overlap between normal range and therapeutic DOAC levels. 18 Low ACT values were reported at baseline in patients under DOAC or VKA with considerable overlap. 19 At supratherapeutic concentrations, ACTs are consistently influenced by DOAC while therapeutic concentrations showed differential effects. 20 ACT is known to efficiently evaluate the anticoagulation state provided by VKA + heparin, which has additive effects. 4  ACTs were significantly lower and heparin requirements were higher in patients under uninterrupted apixaban than VKA in another work. 16 One meta-analysis also showed that use of VKA was associated with reduced heparin requirements and with lower time to achieve the target ACT compared to DOAC either interrupted or uninterrupted. 17 No control group does seem to have been ever investigated before. Our study is therefore the first one to compare ACT and heparin requirements between DOAC and controls, demonstrating that heparin requirements are similar for achieving similar ACT in patients with DOAC.

| D ISCUSS I ON
We did not find any significant difference between rivaroxaban and apixaban for any coagulation time or heparin dosing throughout the procedure. Concentrations of rivaroxaban have been suspected to be better correlated to ACT than apixaban with or without heparin, 7,22,23 and rivaroxaban achieved better coagulation control and apixaban needed higher heparin requirements in the retrospective japanese study. 12 Other works are needed to confirm or infirm these differences.
Potential solutions for performing ablation procedures with a more reliable coagulation control would be to switch patients to VKA, to interrupt DOAC (but this is not advised by the guidelines),

| LI M ITATI O N S
We did not measure baseline ACT. This would have maybe demonstrated baseline differences between groups, 12,19 possibly explaining some of the differences noted thereafter during the procedure.
However baseline ACT has been shown to be in the normal range with slight differences between some DOAC and VKA. 19 Because more rarely prescribed in our area, dabigatran was not used in this study. Thus no speculation can be made for this drug.

CO N FLI C T O F I NTE R E S T
Authors declare no conflict of interest for this article.

Philippe Maury
https://orcid.org/0000-0002-4800-1841 F I G U R E 1 upper: evolution in activated clotting time (ACT) throughout the procedure in each group. ACT> 400 were coded as being measured for 450 seconds for graphical purposes. Lower: evolution in heparin requirements throughout the procedure in each group. Discontinuous line in vitamin K antagonists patients is explained by the lack of heparin need in any patient at some times