ASAIO Journal 2014
Adult Circulatory Support
Safety of Anticoagulation Reversal in Patients Supported with Continuous-Flow Left Ventricular Assist Devices Douglas L. Jennings,* Minu Jacob,† Anuvrat Chopra,† Carrie W. Nemerovski,‡ Jeffrey A. Morgan,§ and David E. Lanfear¶
C
The purpose of this study was to characterize the potential thromboembolic risk associated with the reversal of w arfarin-based anticoagulation in patients with continuous-flow left ventricular assist devices (CF-LVADs). All patients implanted with a CF-LVAD between January 1, 2008, and August 1, 2012, at our institution were screened, and those who received anticoagulation reversal during an inpatient admission were enrolled. The primary outcome is the incidence of thrombotic events, including stroke, device thrombosis, or venous thromboembolism within 30 days of anticoagulation reversal. Of the 122 patients screened, 25 patients experienced 38 anticoagulation reversal events. All patients received vitamin K at a mean dose of 10 ± 8 mg, while 60% of patients received fresh frozen plasma. Only two patients received prothombin complex concentration and three patients received activated factor VII. The rate of thromboembolism within 30 days of attempted reversal was 2.6% (1/38). This patient developed an ischemic stroke after reversal with a high dose of activated factor VII for an acute intracranial bleed. The mortality rate within 30 days of reversal was 20% (5/25), with three of these deaths associated with acute intracranial hemorrhage. Anticoagulation reversal may be safely attempted in selected patients under CF-LVAD support. ASAIO Journal 2014; 60:381–384.
ontinuous-flow left ventricular assist devices (CF-LVADs) represent a lifesaving modality for patients afflicted with e nd-stage heart failure.1 Unfortunately, device-related complications are frequent, with fewer than 30% of patients remaining free of major adverse events during the first year of support.2 Bleeding, particularly from the gastrointestinal (GI) tract, is a common complication associated with CF-LVADs.3 This is problematic as these patients require long-term anticoagulation to prevent thromboembolic events, such as stroke and device-related thrombosis. Guidelines from the American College of Chest Physicians provide recommendations for the reversal of warfarin-based anticoagulation.4 In general, these guidelines advocate that low-dose oral vitamin K can be considered for significant coagulopathy without bleeding, whereas high-dose intravenous (IV) vitamin K along with either fresh-frozen plasma (FFP) or prothrombin complex concentrate (PCC) should be used to manage life-threatening bleeding. However, there are no data presently available to quantify the thromboembolic risk associated with anticoagulation reversal in patients with CF-LVADs. The purpose of this study was to examine the safety of reversing warfarin-based anticoagulation in CF-LVAD patients. Methods For this retrospective cohort study, all patients implanted with a CF-LVAD between January 1, 2008, and August 1, 2012, were screened for inclusion. The screening process began by querying the pharmacy system database to identify physician orders for vitamin K. Those patients who were confirmed to have received anticoagulation reversal during an inpatient admission were enrolled. Patients with multiple reversal attempts could be enrolled twice, provided at least 30 days had elapsed between encounters. At our institution, all patients receive antiplatelet therapy with aspirin 81–325 mg daily, along with warfarin. The goal international normalized ratio (INR) is determined by the type of device, along with patient-specific factors (i.e., atrial fibrillation), and is usually 2–3. Data collection included patient demographics, type of device, goal of device support (bridge to transplant versus destination therapy), indication for reversal, location of hemorrhage (when applicable), type of procedure (when applicable), type and dose of the anticoagulation reversal agent, and the use of parenteral anticoagulant for bridging therapy after the reversal attempt. The primary outcome is the incidence of thrombotic events, including stroke, device thrombosis, or venous thromboembolism (i.e., deep vein thrombosis or pulmonary embolism)
Key Words: anticoagulation, reversal, bleeding, left ventricular assist devices From the *Department of Pharmacy Practice, Nova Southeastern University, Ft. Lauderdale, Florida; †Wayne State University, Detroit, Michigan; ‡Department of Pharmacy Services, Henry Ford Hospital, Detroit, Michigan; §Department of Cardiac Surgery, Henry Ford Hospital, Detroit, Michigan; and ¶Department of Cardiovascular Medicine, Henry Ford Hospital, Detroit, Michigan. Submitted for consideration September 2013; accepted for publication in revised form February 2014. Disclosure: J. A. Morgan has research relationships with HeartWare. D. E. Lanfear has research relationships with Thoratec and HeartWare. He has also served as a speaker for Thoratec. No external funding was used to complete this project. All other authors declare no conflict of interest. Correspondence: Douglas L. Jennings, PharmD, Nova Southeastern University College of Pharmacy 3200 South University Drive Fort Lauderdale, FL 33328. Email: [email protected]. Copyright © 2014 by the American Society for Artificial Internal Organs DOI: 10.1097/MAT.0000000000000077
381
382 JENNINGS et al. within 30 days of anticoagulation reversal. Additional data including the INR and hemoglobin values before and after reversal, as well as 30 day mortality, were also analyzed. Data are presented as mean ± SD, percentage, or frequency distribution, depending on what is most appropriate. All patients who met inclusion criteria were analyzed. Patient demographics and clinical endpoints were characterized with descriptive statistics using Microsoft Excel. Results During the screening period, 122 patients were implanted with CF-LVADs at our institution. Of these, 25 patients experienced 38 anticoagulation reversal events. Baseline demographic data are presented in Table 1. The rate of thromboembolism within 30 days of attempted reversal was 2.6% (1/38). The mortality rate within 30 days of reversal was 20% (5/25); causes of death were attributed to intracranial hemorrhage (3), sepsis with Bacteroides bacteremia (1), and pulmonary aspergillus infection (1). The indication for reversal was acute hemorrhage (19), invasive procedure (7), supratherapeutic INR (9), and unexplained anemia (3). Table 2 summarizes data on the use of reversal agents in the overall population. Reversal for Acute Hemorrhage Acute gastrointestinal bleeding. Six patients had a total of eight reversal attempts for GI hemorrhage. Hemoglobin and INR values before and after reversal are displayed in Table 3. Details on reversal agents are presented in Table 4. There were no fatalities or thromboembolic events within 30 days, and all patients were restarted on anticoagulation upon discharge. In six attempts, warfarin was restarted without parenteral bridging; IV heparin was used in the remaining two cases. Acute intracranial bleeding. Five patients had five reversal attempts for acute intracranial bleeding. Hemoglobin and INR values before and after reversal are displayed in Table 3. Details Table 1. Baseline Patient Characteristics (N = 25) Characteristic Age, y, mean ± SD Gender (male) Race White African American Unknown/other Etiology Nonischemic Ischemic Device HeartMate II HeartWare HVAD Indication Bridge to transplant Bridge to destination Hypertension Chronic kidney disease (estimated GFR < 60 ml/min) Diabetes mellitus Atrial fibrillation
Percentage (n) 53 ± 13 84 (21) 28 (7) 64 (16) 8 (2)
on the specific reversal agents used are presented in Tables 2 and 4. There were three fatalities within 30 days of reversal in this group, all resulting from the sequelae of the intracranial hemorrhage. There was one thromboembolic complication within 30 days of reversal; this patient received 10 mg of activated factor VII, 9 units of FFP, and a total of 30 mg of IV vitamin K during his reversal attempt, which occurred approximately 185 days after initial implant. On hospital day 15, the patient became less responsive with a newly fixed left pupil, and repeat imaging demonstrated a new ischemic stroke in the posterior right basal ganglia. Given the clinical deterioration and poor prognosis, the decision was made to withdraw support on hospital day 17. Miscellaneous bleeding. Six patients had six reversal attempts for various bleeding episodes, including hematoma (3), epistaxis (2), and intra-abdominal bleed (1). Hemoglobin and INR values for this cohort before and after reversal are displayed in Table 3. Details on the reversal agents used are presented in Table 4. There were no thromboembolic events or fatalities within 30 days. All patients were restarted on anticoagulation upon discharge, and no patients received parenteral bridging after reversal. Reversal for Invasive Procedures Seven patients had seven reversal attempts for invasive procedures, which included thoracentesis (4), elective surgery (2), and wound debridement (1). Hemoglobin and INR values before and after reversal are displayed in Table 3. Details on the reversal agents used are presented in Table 4. There were no thromboembolic events within 30 days, while one patient died in home hospice secondary pulmonary aspergillus infection approximately 3 weeks after his reversal event. All patients were restarted on anticoagulation upon discharge. Two patients were bridged with enoxaparin, whereas one patient received IV heparin. Reversal for Elevated INR Values Seven patients had nine reversal attempts for elevated INR values. Hemoglobin and INR values before and after reversal are displayed in Table 3. Information about the reversal agents used is presented in Table 4. There were no thromboembolic events within 30 days, while there was one fatality within 30 days (sepsis). All patients who survived were restarted on anticoagulation upon discharge. No parenteral bridging was used in this group. Reversal for Anemia
56 (14) 44 (11) 88 (22) 44 (11)
Three patients received anticoagulation reversal for acute anemia without overt evidence of hemorrhage. These patients were reversed based on the suspicion for bleeding; however, upon further diagnostic testing none was located. Hemoglobin and INR values before and after reversal are displayed in Table 3. Details on the reversal agents used are presented in Table 4. There were no thromboembolic events or fatalities within 30 days. All patients were restarted on anticoagulation upon discharge, and no parenteral bridging was used.
36 (9) 50 (13)
Discussion
72 (18) 28 (7) 80 (20) 20 (5)
GFR, glomerular filtration rate; HVAD, HeartWare ventricular assist device.
The cornerstone of thromboprophylaxis in patients with contemporary CF-LVADs remains warfarin-based anticoagulation.
383
ANTICOAGULATION REVERSAL IN CF-LVAD PATIENTS Table 2. Use of Specific Reversal Agents (N = 38)
Agent Vitamin K
Incident Use (n) 100% (38)
FFP pRBC PCC* Activated factor VII*
60% (23) 45% (17) 5% (2) 8% (3)
Parenteral anticoagulation bridging after reversal
16% (6)
Mean Dose (±SD) Route (n) 10 ± 8 mg Oral: 47% (18) IV: 42% (16) Oral and IV: 13% (5) Subcutaneous: 2% (1) 4.4 ± 3 units 4.6 ± 3.8 units 1,500 units per patient Patient 1: 1 mg Patient 2: 1.7 mg Patient 3: 10 mg Heparin†: 4 Enoxaparin‡: 2
*These agents were only used in patients with acute intracranial bleeding. †Heparin was dosed to achieve a goal activated partial thromboplastin time between 55 and 75 seconds. ‡Enoxaparin dosed 1 mg/kg every 12 hours (adjusted for renal function). FFP, fresh-frozen plasma; IV, intravenous; PCC, three-factor prothrombin complex concentrate; pRBC, packed red blood cells.
This mandatory long-term anticoagulation is problematic as support from these devices increases the risk of bleeding through depletion of von Willebrand factor and loss of pulsatility.5,6 Bleeding in this population represents a leading cause of hospital readmission,7 and data from the most recent Interagency Registry for Mechanically Assisted Circulatory Support report suggest that the rate of hemorrhage during the first year of support is nearly 10 events per 100 patient-months of follow-up. Strategies to reverse the effect of warfarin may be needed in those who require an urgent invasive procedure, in asymptomatic patients presenting with excessively elevated INR values, or in those with acute hemorrhage. Current guidelines recommend that interruption of warfarin or low-dose oral vitamin K may be sufficient for those undergoing a procedure or who have asymptomatic INR elevations.4 In general, for patients receiving warfarin, it takes approximately 2.5 days for an INR between 6 and 10 to decline to
Adult Circulatory Support
Safety of Anticoagulation Reversal in Patients Supported with Continuous-Flow Left Ventricular Assist Devices Douglas L. Jennings,* Minu Jacob,† Anuvrat Chopra,† Carrie W. Nemerovski,‡ Jeffrey A. Morgan,§ and David E. Lanfear¶
C
The purpose of this study was to characterize the potential thromboembolic risk associated with the reversal of w arfarin-based anticoagulation in patients with continuous-flow left ventricular assist devices (CF-LVADs). All patients implanted with a CF-LVAD between January 1, 2008, and August 1, 2012, at our institution were screened, and those who received anticoagulation reversal during an inpatient admission were enrolled. The primary outcome is the incidence of thrombotic events, including stroke, device thrombosis, or venous thromboembolism within 30 days of anticoagulation reversal. Of the 122 patients screened, 25 patients experienced 38 anticoagulation reversal events. All patients received vitamin K at a mean dose of 10 ± 8 mg, while 60% of patients received fresh frozen plasma. Only two patients received prothombin complex concentration and three patients received activated factor VII. The rate of thromboembolism within 30 days of attempted reversal was 2.6% (1/38). This patient developed an ischemic stroke after reversal with a high dose of activated factor VII for an acute intracranial bleed. The mortality rate within 30 days of reversal was 20% (5/25), with three of these deaths associated with acute intracranial hemorrhage. Anticoagulation reversal may be safely attempted in selected patients under CF-LVAD support. ASAIO Journal 2014; 60:381–384.
ontinuous-flow left ventricular assist devices (CF-LVADs) represent a lifesaving modality for patients afflicted with e nd-stage heart failure.1 Unfortunately, device-related complications are frequent, with fewer than 30% of patients remaining free of major adverse events during the first year of support.2 Bleeding, particularly from the gastrointestinal (GI) tract, is a common complication associated with CF-LVADs.3 This is problematic as these patients require long-term anticoagulation to prevent thromboembolic events, such as stroke and device-related thrombosis. Guidelines from the American College of Chest Physicians provide recommendations for the reversal of warfarin-based anticoagulation.4 In general, these guidelines advocate that low-dose oral vitamin K can be considered for significant coagulopathy without bleeding, whereas high-dose intravenous (IV) vitamin K along with either fresh-frozen plasma (FFP) or prothrombin complex concentrate (PCC) should be used to manage life-threatening bleeding. However, there are no data presently available to quantify the thromboembolic risk associated with anticoagulation reversal in patients with CF-LVADs. The purpose of this study was to examine the safety of reversing warfarin-based anticoagulation in CF-LVAD patients. Methods For this retrospective cohort study, all patients implanted with a CF-LVAD between January 1, 2008, and August 1, 2012, were screened for inclusion. The screening process began by querying the pharmacy system database to identify physician orders for vitamin K. Those patients who were confirmed to have received anticoagulation reversal during an inpatient admission were enrolled. Patients with multiple reversal attempts could be enrolled twice, provided at least 30 days had elapsed between encounters. At our institution, all patients receive antiplatelet therapy with aspirin 81–325 mg daily, along with warfarin. The goal international normalized ratio (INR) is determined by the type of device, along with patient-specific factors (i.e., atrial fibrillation), and is usually 2–3. Data collection included patient demographics, type of device, goal of device support (bridge to transplant versus destination therapy), indication for reversal, location of hemorrhage (when applicable), type of procedure (when applicable), type and dose of the anticoagulation reversal agent, and the use of parenteral anticoagulant for bridging therapy after the reversal attempt. The primary outcome is the incidence of thrombotic events, including stroke, device thrombosis, or venous thromboembolism (i.e., deep vein thrombosis or pulmonary embolism)
Key Words: anticoagulation, reversal, bleeding, left ventricular assist devices From the *Department of Pharmacy Practice, Nova Southeastern University, Ft. Lauderdale, Florida; †Wayne State University, Detroit, Michigan; ‡Department of Pharmacy Services, Henry Ford Hospital, Detroit, Michigan; §Department of Cardiac Surgery, Henry Ford Hospital, Detroit, Michigan; and ¶Department of Cardiovascular Medicine, Henry Ford Hospital, Detroit, Michigan. Submitted for consideration September 2013; accepted for publication in revised form February 2014. Disclosure: J. A. Morgan has research relationships with HeartWare. D. E. Lanfear has research relationships with Thoratec and HeartWare. He has also served as a speaker for Thoratec. No external funding was used to complete this project. All other authors declare no conflict of interest. Correspondence: Douglas L. Jennings, PharmD, Nova Southeastern University College of Pharmacy 3200 South University Drive Fort Lauderdale, FL 33328. Email: [email protected]. Copyright © 2014 by the American Society for Artificial Internal Organs DOI: 10.1097/MAT.0000000000000077
381
382 JENNINGS et al. within 30 days of anticoagulation reversal. Additional data including the INR and hemoglobin values before and after reversal, as well as 30 day mortality, were also analyzed. Data are presented as mean ± SD, percentage, or frequency distribution, depending on what is most appropriate. All patients who met inclusion criteria were analyzed. Patient demographics and clinical endpoints were characterized with descriptive statistics using Microsoft Excel. Results During the screening period, 122 patients were implanted with CF-LVADs at our institution. Of these, 25 patients experienced 38 anticoagulation reversal events. Baseline demographic data are presented in Table 1. The rate of thromboembolism within 30 days of attempted reversal was 2.6% (1/38). The mortality rate within 30 days of reversal was 20% (5/25); causes of death were attributed to intracranial hemorrhage (3), sepsis with Bacteroides bacteremia (1), and pulmonary aspergillus infection (1). The indication for reversal was acute hemorrhage (19), invasive procedure (7), supratherapeutic INR (9), and unexplained anemia (3). Table 2 summarizes data on the use of reversal agents in the overall population. Reversal for Acute Hemorrhage Acute gastrointestinal bleeding. Six patients had a total of eight reversal attempts for GI hemorrhage. Hemoglobin and INR values before and after reversal are displayed in Table 3. Details on reversal agents are presented in Table 4. There were no fatalities or thromboembolic events within 30 days, and all patients were restarted on anticoagulation upon discharge. In six attempts, warfarin was restarted without parenteral bridging; IV heparin was used in the remaining two cases. Acute intracranial bleeding. Five patients had five reversal attempts for acute intracranial bleeding. Hemoglobin and INR values before and after reversal are displayed in Table 3. Details Table 1. Baseline Patient Characteristics (N = 25) Characteristic Age, y, mean ± SD Gender (male) Race White African American Unknown/other Etiology Nonischemic Ischemic Device HeartMate II HeartWare HVAD Indication Bridge to transplant Bridge to destination Hypertension Chronic kidney disease (estimated GFR < 60 ml/min) Diabetes mellitus Atrial fibrillation
Percentage (n) 53 ± 13 84 (21) 28 (7) 64 (16) 8 (2)
on the specific reversal agents used are presented in Tables 2 and 4. There were three fatalities within 30 days of reversal in this group, all resulting from the sequelae of the intracranial hemorrhage. There was one thromboembolic complication within 30 days of reversal; this patient received 10 mg of activated factor VII, 9 units of FFP, and a total of 30 mg of IV vitamin K during his reversal attempt, which occurred approximately 185 days after initial implant. On hospital day 15, the patient became less responsive with a newly fixed left pupil, and repeat imaging demonstrated a new ischemic stroke in the posterior right basal ganglia. Given the clinical deterioration and poor prognosis, the decision was made to withdraw support on hospital day 17. Miscellaneous bleeding. Six patients had six reversal attempts for various bleeding episodes, including hematoma (3), epistaxis (2), and intra-abdominal bleed (1). Hemoglobin and INR values for this cohort before and after reversal are displayed in Table 3. Details on the reversal agents used are presented in Table 4. There were no thromboembolic events or fatalities within 30 days. All patients were restarted on anticoagulation upon discharge, and no patients received parenteral bridging after reversal. Reversal for Invasive Procedures Seven patients had seven reversal attempts for invasive procedures, which included thoracentesis (4), elective surgery (2), and wound debridement (1). Hemoglobin and INR values before and after reversal are displayed in Table 3. Details on the reversal agents used are presented in Table 4. There were no thromboembolic events within 30 days, while one patient died in home hospice secondary pulmonary aspergillus infection approximately 3 weeks after his reversal event. All patients were restarted on anticoagulation upon discharge. Two patients were bridged with enoxaparin, whereas one patient received IV heparin. Reversal for Elevated INR Values Seven patients had nine reversal attempts for elevated INR values. Hemoglobin and INR values before and after reversal are displayed in Table 3. Information about the reversal agents used is presented in Table 4. There were no thromboembolic events within 30 days, while there was one fatality within 30 days (sepsis). All patients who survived were restarted on anticoagulation upon discharge. No parenteral bridging was used in this group. Reversal for Anemia
56 (14) 44 (11) 88 (22) 44 (11)
Three patients received anticoagulation reversal for acute anemia without overt evidence of hemorrhage. These patients were reversed based on the suspicion for bleeding; however, upon further diagnostic testing none was located. Hemoglobin and INR values before and after reversal are displayed in Table 3. Details on the reversal agents used are presented in Table 4. There were no thromboembolic events or fatalities within 30 days. All patients were restarted on anticoagulation upon discharge, and no parenteral bridging was used.
36 (9) 50 (13)
Discussion
72 (18) 28 (7) 80 (20) 20 (5)
GFR, glomerular filtration rate; HVAD, HeartWare ventricular assist device.
The cornerstone of thromboprophylaxis in patients with contemporary CF-LVADs remains warfarin-based anticoagulation.
383
ANTICOAGULATION REVERSAL IN CF-LVAD PATIENTS Table 2. Use of Specific Reversal Agents (N = 38)
Agent Vitamin K
Incident Use (n) 100% (38)
FFP pRBC PCC* Activated factor VII*
60% (23) 45% (17) 5% (2) 8% (3)
Parenteral anticoagulation bridging after reversal
16% (6)
Mean Dose (±SD) Route (n) 10 ± 8 mg Oral: 47% (18) IV: 42% (16) Oral and IV: 13% (5) Subcutaneous: 2% (1) 4.4 ± 3 units 4.6 ± 3.8 units 1,500 units per patient Patient 1: 1 mg Patient 2: 1.7 mg Patient 3: 10 mg Heparin†: 4 Enoxaparin‡: 2
*These agents were only used in patients with acute intracranial bleeding. †Heparin was dosed to achieve a goal activated partial thromboplastin time between 55 and 75 seconds. ‡Enoxaparin dosed 1 mg/kg every 12 hours (adjusted for renal function). FFP, fresh-frozen plasma; IV, intravenous; PCC, three-factor prothrombin complex concentrate; pRBC, packed red blood cells.
This mandatory long-term anticoagulation is problematic as support from these devices increases the risk of bleeding through depletion of von Willebrand factor and loss of pulsatility.5,6 Bleeding in this population represents a leading cause of hospital readmission,7 and data from the most recent Interagency Registry for Mechanically Assisted Circulatory Support report suggest that the rate of hemorrhage during the first year of support is nearly 10 events per 100 patient-months of follow-up. Strategies to reverse the effect of warfarin may be needed in those who require an urgent invasive procedure, in asymptomatic patients presenting with excessively elevated INR values, or in those with acute hemorrhage. Current guidelines recommend that interruption of warfarin or low-dose oral vitamin K may be sufficient for those undergoing a procedure or who have asymptomatic INR elevations.4 In general, for patients receiving warfarin, it takes approximately 2.5 days for an INR between 6 and 10 to decline to
