J Interv Card Electrophysiol (2014) 41:231–236 DOI 10.1007/s10840-014-9948-1
A comparison of bleeding complications between warfarin, dabigatran, and rivaroxaban in patients undergoing cryoballoon ablation David Snipelisky & Jordan C. Ray & Ryan Ung & Melissa Duart & Christine Kauffman & Fred Kusumoto
Received: 16 July 2014 / Accepted: 1 September 2014 / Published online: 8 November 2014 # Springer Science+Business Media New York 2014
Abstract Introduction In recent years, several novel anticoagulants have been approved for the prevention of thromboembolic strokes as an alternative to warfarin in patients with atrial arrhythmias. Studies have evaluated these medications in patients undergoing radiofrequency ablation, yet no data exists to evaluate the bleeding risk in patients undergoing cryoballoon ablation procedures. Methods Patients that underwent either cryoballoon ablation alone or with additional radiofrequency ablation over the past 3 years were included in the study. Patients were stratified into one of three subsets based on type of anticoagulation (warfarin, dabigatran, or rivaroxaban). Bleeding complications during the first 48 h and first 2 weeks following the ablation were recorded. Major complications were defined as hemorrhage requiring blood products or need for vascular intervention. Minor complications included prolonged bleeding from catheter insertion site, development of ecchymosis, or hematoma formation. Intraprocedural activated clotting times (ACT) were assessed and compared. Results A total of 217 patients met inclusion criteria of which 87 (40.1 %) patients were on warfarin, 90 (41.5 %) patients on dabigatran, and 40 (18.4 %) patients on rivaroxaban. The overall bleeding complication rate was 12.0 %. All complications occurred within the first 48 h post-ablation. Nine (10.3 %) complications occurred in the warfarin subset, ten (11.1 %) in the rivaroxaban subset, and seven (17.5 %) in the
D. Snipelisky (*) : J. C. Ray : R. Ung : M. Duart Department of Medicine, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL 32224, USA e-mail: [email protected] C. Kauffman : F. Kusumoto Department of Medicine, Division of Cardiovascular Diseases, Jacksonville, FL 32224, USA
dabigatran subset (p=0.49). The warfarin and dabigatran subsets had higher average ACT levels (424.9 versus 406.5) compared to the rivaroxaban subset (393.4; p < 0.01). Subanalyses found no difference in bleeding complications based on procedure type. Conclusion Bleeding complications post-ablation were similar for warfarin, dabigatran, and rivaroxaban in patients undergoing cryoballoon ablation. Compared with radiofrequency ablation, cryoablation does not place patients at an increased bleeding risk. Keywords Novel anticoagulants . Bleeding risk . Cryoablation . Radiofrequency ablation
1 Introduction Radiofrequency ablation has been an important treatment option for selected patients with symptomatic atrial fibrillation that have failed traditional therapies [1–4]. Over the last several years, cryoballoon ablation of the pulmonary vein antrum has become an increasingly popular technique. In addition, combined hybrid procedures, incorporating both cryo- and radiofrequency ablative techniques, have also become more widely used, especially in those patients who had failed prior ablative studies and have multiple foci for atrial tachycardia [3, 4]. Although the efficacy of both procedures is well-established, one of the potential downsides of cryoballoon ablation with or without additional radiofrequency ablation is the larger overall sheath size and increased number of sheaths used, potentiating the risk of bleeding [1, 2]. Although the novel anticoagulants dabigatran and rivaroxaban have demonstrated non-inferiority to warfarin for stroke prevention in atrial fibrillation and studies have shown that the novel anticoagulants do not increase bleeding
232
risk following radiofrequency ablative procedures, little data is present regarding bleeding risks following cryoballoon ablation or more extensive ablation procedures that combine cryoballoon ablation and radiofrequency ablation [5–14]. The aim of our study is to compare the incidence of bleeding associated with warfarin and the novel anticoagulants dabigatran and rivaroxaban in patients undergoing cryoballoon ablation with or without additional radiofrequency ablation.
2 Methods Consecutive patients that underwent cryoballoon ablation of atrial fibrillation at our institution over the past 3 years were evaluated via retrospective chart review. Patients were stratified into one of three subsets based on type of anticoagulant— warfarin, dabigatran, or rivaroxaban. Clinical documentation, including procedure note, nurse notes, physician progress notes, and clinic or hospital visit within 2 weeks of the procedure were reviewed. All laboratory data collected was within 30 days prior to the procedure in a non-emergent setting. Data were collected via the electronic medical record. Demographic and clinical characteristics, including CHADS2 score, presence of renal or liver dysfunction, alcohol use, concomitant use of antiplatelet therapy and, in the warfarin subset, international normalization ratio (INR) were evaluated. Patients were defined as having renal dysfunction if the creatinine was greater than 2.0 mg/dL or if the patient required dialysis. Liver dysfunction was defined as documented chronic hepatitis or biochemical evidence of significant hepatic dysfunction illustrated by alanine aminotransferase or aspartate aminotransferase levels greater than three times normal or total bilirubin greater than twice normal. Consumption of more than eight alcoholic beverages weekly characterized a patient as an alcohol user. HAS-BLED and HEMORR2HAGES scores were calculated based on data available at time of ablation. Complications during the procedure and at 48 h and 2 week time periods were assessed. Minor complications were defined as prolonged bleeding from the catheter insertion site requiring re-application of the femoral artery compression device, hematoma formation, or presence of ecchymosis following the procedure. Major complications included hemorrhage requiring vascular intervention or the need for blood products [7–9]. Activated clotting times (ACT) were collected on each patient. Average, minimum, and maximum ACT levels were calculated as well as average ACT variation between maximum and minimum values within each subset. Intraprocedure ACT was managed via institutional protocol. An initial loading dose of 8000 to 12,000 units of heparin was given, followed by subsequent doses of 1000 to 4000 units
J Interv Card Electrophysiol (2014) 41:231–236
every 15 min with a goal to achieve an average ACT of at least 350 s. Seldinger technique without ultrasound guidance was used to establish vascular access at procedure initiation and sheaths were removed immediately following procedure completion prior to patient transfer to the recovery center. Sheaths were placed in both femoral veins, and three sheaths were primarily used, although in patients with more complex arrhythmias, one additional mapping sheath was placed. Protamine sulfate was not routinely used following the procedure. Patients with incomplete documentation or lapses in laboratory evaluations in the medical record were excluded. All patients were on one type of anticoagulant both preceding and following ablation. No patient switched from one type to another within 3 months of the procedure; otherwise, they were excluded in the study analysis. All patients resumed the respective anticoagulant the evening following ablation. Patients on warfarin and rivaroxaban were instructed to hold the medication dose the evening prior to the procedure, while patients on dabigatran held the dose the morning of the procedure. Patients were not bridged with low molecular weight heparin products prior to the procedure. Patients on warfarin therapy had a recorded INR of between 2.0 to 3.0 within 14 days of the ablation. Lack of physical examination or incomplete medication list excluded a patient from the study. Two-tailed Fisher’s exact test and analysis of variance calculations were used for statistical analysis. P values of less than 0.05 were considered significant. Comparisons between demographic, clinical, and laboratory factors as well as bleeding complications at both the 48 h and 2 week time periods were made. Institutional review board approval was obtained prior to study initiation.
3 Results Over the past 3 years, a total of 487 patients underwent ablation for atrial fibrillation and 217 underwent cryoballoon ablation and met inclusion criteria and were included in the main analysis. Of these patients, 87 (40.1 %) were anticoagulated with warfarin, 90 (41.5 %) with dabigatran, and 40 (18.4 %) with rivaroxaban. No significant differences were found in age, sex, body mass index, and presence of hepatic dysfunction between patient subsets. Patients in the warfarin subset had a higher overall CHADS2 score of 1.4 compared with 1.04 in the dabigatran subset and 0.95 in the rivaroxaban subset (p=0.02). A greater number of patients had a history of congestive heart failure in the warfarin subset (n=11, 12.6 %) when compared with the dabigatran (n=3, 3.3 %) and rivaroxaban (n=2, 5.0 %) subsets (p=0.05). Similarly, higher proportions of patients had a history of diabetes
J Interv Card Electrophysiol (2014) 41:231–236
mellitus (p=0.08) and abnormal renal function (p=0.021) in the warfarin subset. No differences were noted in the presence of stroke/transient ischemic attack (p=0.488) and hypertension (p=0.23). Nine (10.3 %) patients in the warfarin subset were on concomitant antiplatelet therapy compared to three patients (3.3 %) and two patients (5.0 %) in the dabigatran and rivaroxaban subsets, respectively (p=0.153). A higher number of patients had a history of alcohol use in the warfarin (n= 18, 20.7 %) and rivaroxaban (n=11, 27.5 %) subsets when compared to the dabigatran (n=10, 11.1 %) subset (p=0.056) (Table 1). The average HAS-BLED score was highest in the warfarin and rivaroxaban subsets with a value of 1.1 compared to 0.71 in the dabigatran subset (p=0.008). The warfarin subset was found to have a slightly higher HEMORR2HAGES score of 0.98 when compared to the rivaroxaban (score of 0.88) and dabigatran (score of 0.62) subsets (p= 0.086). No embolic events were recorded within the study period in patients on either warfarin, dabigatran, or rivaroxaban. Procedure time was similar between subsets (p=0.096). The average intraprocedure ACT level was lowest in the rivaroxaban subset (393.4 s) compared to the warfarin (424.9 s) and dabigatran (406.5 s) subsets (p
A comparison of bleeding complications between warfarin, dabigatran, and rivaroxaban in patients undergoing cryoballoon ablation David Snipelisky & Jordan C. Ray & Ryan Ung & Melissa Duart & Christine Kauffman & Fred Kusumoto
Received: 16 July 2014 / Accepted: 1 September 2014 / Published online: 8 November 2014 # Springer Science+Business Media New York 2014
Abstract Introduction In recent years, several novel anticoagulants have been approved for the prevention of thromboembolic strokes as an alternative to warfarin in patients with atrial arrhythmias. Studies have evaluated these medications in patients undergoing radiofrequency ablation, yet no data exists to evaluate the bleeding risk in patients undergoing cryoballoon ablation procedures. Methods Patients that underwent either cryoballoon ablation alone or with additional radiofrequency ablation over the past 3 years were included in the study. Patients were stratified into one of three subsets based on type of anticoagulation (warfarin, dabigatran, or rivaroxaban). Bleeding complications during the first 48 h and first 2 weeks following the ablation were recorded. Major complications were defined as hemorrhage requiring blood products or need for vascular intervention. Minor complications included prolonged bleeding from catheter insertion site, development of ecchymosis, or hematoma formation. Intraprocedural activated clotting times (ACT) were assessed and compared. Results A total of 217 patients met inclusion criteria of which 87 (40.1 %) patients were on warfarin, 90 (41.5 %) patients on dabigatran, and 40 (18.4 %) patients on rivaroxaban. The overall bleeding complication rate was 12.0 %. All complications occurred within the first 48 h post-ablation. Nine (10.3 %) complications occurred in the warfarin subset, ten (11.1 %) in the rivaroxaban subset, and seven (17.5 %) in the
D. Snipelisky (*) : J. C. Ray : R. Ung : M. Duart Department of Medicine, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL 32224, USA e-mail: [email protected] C. Kauffman : F. Kusumoto Department of Medicine, Division of Cardiovascular Diseases, Jacksonville, FL 32224, USA
dabigatran subset (p=0.49). The warfarin and dabigatran subsets had higher average ACT levels (424.9 versus 406.5) compared to the rivaroxaban subset (393.4; p < 0.01). Subanalyses found no difference in bleeding complications based on procedure type. Conclusion Bleeding complications post-ablation were similar for warfarin, dabigatran, and rivaroxaban in patients undergoing cryoballoon ablation. Compared with radiofrequency ablation, cryoablation does not place patients at an increased bleeding risk. Keywords Novel anticoagulants . Bleeding risk . Cryoablation . Radiofrequency ablation
1 Introduction Radiofrequency ablation has been an important treatment option for selected patients with symptomatic atrial fibrillation that have failed traditional therapies [1–4]. Over the last several years, cryoballoon ablation of the pulmonary vein antrum has become an increasingly popular technique. In addition, combined hybrid procedures, incorporating both cryo- and radiofrequency ablative techniques, have also become more widely used, especially in those patients who had failed prior ablative studies and have multiple foci for atrial tachycardia [3, 4]. Although the efficacy of both procedures is well-established, one of the potential downsides of cryoballoon ablation with or without additional radiofrequency ablation is the larger overall sheath size and increased number of sheaths used, potentiating the risk of bleeding [1, 2]. Although the novel anticoagulants dabigatran and rivaroxaban have demonstrated non-inferiority to warfarin for stroke prevention in atrial fibrillation and studies have shown that the novel anticoagulants do not increase bleeding
232
risk following radiofrequency ablative procedures, little data is present regarding bleeding risks following cryoballoon ablation or more extensive ablation procedures that combine cryoballoon ablation and radiofrequency ablation [5–14]. The aim of our study is to compare the incidence of bleeding associated with warfarin and the novel anticoagulants dabigatran and rivaroxaban in patients undergoing cryoballoon ablation with or without additional radiofrequency ablation.
2 Methods Consecutive patients that underwent cryoballoon ablation of atrial fibrillation at our institution over the past 3 years were evaluated via retrospective chart review. Patients were stratified into one of three subsets based on type of anticoagulant— warfarin, dabigatran, or rivaroxaban. Clinical documentation, including procedure note, nurse notes, physician progress notes, and clinic or hospital visit within 2 weeks of the procedure were reviewed. All laboratory data collected was within 30 days prior to the procedure in a non-emergent setting. Data were collected via the electronic medical record. Demographic and clinical characteristics, including CHADS2 score, presence of renal or liver dysfunction, alcohol use, concomitant use of antiplatelet therapy and, in the warfarin subset, international normalization ratio (INR) were evaluated. Patients were defined as having renal dysfunction if the creatinine was greater than 2.0 mg/dL or if the patient required dialysis. Liver dysfunction was defined as documented chronic hepatitis or biochemical evidence of significant hepatic dysfunction illustrated by alanine aminotransferase or aspartate aminotransferase levels greater than three times normal or total bilirubin greater than twice normal. Consumption of more than eight alcoholic beverages weekly characterized a patient as an alcohol user. HAS-BLED and HEMORR2HAGES scores were calculated based on data available at time of ablation. Complications during the procedure and at 48 h and 2 week time periods were assessed. Minor complications were defined as prolonged bleeding from the catheter insertion site requiring re-application of the femoral artery compression device, hematoma formation, or presence of ecchymosis following the procedure. Major complications included hemorrhage requiring vascular intervention or the need for blood products [7–9]. Activated clotting times (ACT) were collected on each patient. Average, minimum, and maximum ACT levels were calculated as well as average ACT variation between maximum and minimum values within each subset. Intraprocedure ACT was managed via institutional protocol. An initial loading dose of 8000 to 12,000 units of heparin was given, followed by subsequent doses of 1000 to 4000 units
J Interv Card Electrophysiol (2014) 41:231–236
every 15 min with a goal to achieve an average ACT of at least 350 s. Seldinger technique without ultrasound guidance was used to establish vascular access at procedure initiation and sheaths were removed immediately following procedure completion prior to patient transfer to the recovery center. Sheaths were placed in both femoral veins, and three sheaths were primarily used, although in patients with more complex arrhythmias, one additional mapping sheath was placed. Protamine sulfate was not routinely used following the procedure. Patients with incomplete documentation or lapses in laboratory evaluations in the medical record were excluded. All patients were on one type of anticoagulant both preceding and following ablation. No patient switched from one type to another within 3 months of the procedure; otherwise, they were excluded in the study analysis. All patients resumed the respective anticoagulant the evening following ablation. Patients on warfarin and rivaroxaban were instructed to hold the medication dose the evening prior to the procedure, while patients on dabigatran held the dose the morning of the procedure. Patients were not bridged with low molecular weight heparin products prior to the procedure. Patients on warfarin therapy had a recorded INR of between 2.0 to 3.0 within 14 days of the ablation. Lack of physical examination or incomplete medication list excluded a patient from the study. Two-tailed Fisher’s exact test and analysis of variance calculations were used for statistical analysis. P values of less than 0.05 were considered significant. Comparisons between demographic, clinical, and laboratory factors as well as bleeding complications at both the 48 h and 2 week time periods were made. Institutional review board approval was obtained prior to study initiation.
3 Results Over the past 3 years, a total of 487 patients underwent ablation for atrial fibrillation and 217 underwent cryoballoon ablation and met inclusion criteria and were included in the main analysis. Of these patients, 87 (40.1 %) were anticoagulated with warfarin, 90 (41.5 %) with dabigatran, and 40 (18.4 %) with rivaroxaban. No significant differences were found in age, sex, body mass index, and presence of hepatic dysfunction between patient subsets. Patients in the warfarin subset had a higher overall CHADS2 score of 1.4 compared with 1.04 in the dabigatran subset and 0.95 in the rivaroxaban subset (p=0.02). A greater number of patients had a history of congestive heart failure in the warfarin subset (n=11, 12.6 %) when compared with the dabigatran (n=3, 3.3 %) and rivaroxaban (n=2, 5.0 %) subsets (p=0.05). Similarly, higher proportions of patients had a history of diabetes
J Interv Card Electrophysiol (2014) 41:231–236
mellitus (p=0.08) and abnormal renal function (p=0.021) in the warfarin subset. No differences were noted in the presence of stroke/transient ischemic attack (p=0.488) and hypertension (p=0.23). Nine (10.3 %) patients in the warfarin subset were on concomitant antiplatelet therapy compared to three patients (3.3 %) and two patients (5.0 %) in the dabigatran and rivaroxaban subsets, respectively (p=0.153). A higher number of patients had a history of alcohol use in the warfarin (n= 18, 20.7 %) and rivaroxaban (n=11, 27.5 %) subsets when compared to the dabigatran (n=10, 11.1 %) subset (p=0.056) (Table 1). The average HAS-BLED score was highest in the warfarin and rivaroxaban subsets with a value of 1.1 compared to 0.71 in the dabigatran subset (p=0.008). The warfarin subset was found to have a slightly higher HEMORR2HAGES score of 0.98 when compared to the rivaroxaban (score of 0.88) and dabigatran (score of 0.62) subsets (p= 0.086). No embolic events were recorded within the study period in patients on either warfarin, dabigatran, or rivaroxaban. Procedure time was similar between subsets (p=0.096). The average intraprocedure ACT level was lowest in the rivaroxaban subset (393.4 s) compared to the warfarin (424.9 s) and dabigatran (406.5 s) subsets (p
