Secondary Logo

Journal Logo

Case Reports

Anticoagulation of Impella with a Bivalirudin Purge Solution

Szymanski, Thomas W.*; Weeks, Phillip A.*; Lee, Yeunju; Kumar, Sachin; Castillo, Brian§; Kar, Biswajit; Gregoric, Igor D.

Author Information
doi: 10.1097/MAT.0000000000001126
  • Free


Cardiogenic shock is associated with substantial morbidity and mortality, with a reported 30-day mortality of more than 40%.1 Therapeutic options include inotropes and vasopressors, but many patients develop cardiogenic shock refractory to pharmacological therapy and require mechanical circulatory support to maintain adequate perfusion.2 Percutaneous ventricular assist devices (VADs), including the intraaortic balloon pump (IABP), Impella (Abiomed Inc., Danvers, MA), and TandemHeart (CardiacAssist Inc., Pittsburgh, PA), are used with the goal to serve as temporary support until recovery, as a bridge to durable left VAD (LVAD) insertion, or as a bridge to heart transplantation.3

The Impella is a nonpulsatile, axial flow percutaneous VAD that provides a greater level of hemodynamic support compared with the IABP, and depending on the model, comparable support to the TandemHeart. The Impella is currently available in two different models, the “CP,” which is capable of delivering a flow of up to 3.6 L/min and the “5.0,” which requires surgical insertion and can deliver a flow of up to 5.0 L/min. The Impella is commonly implanted via the femoral artery and is positioned across the aortic valve. The rotor pumps blood from the left ventricle into the ascending aorta, thereby unloading the left ventricle and reducing myocardial oxygen consumption and demand.4–6 A dextrose-based purge solution flows countercurrent to the blood flow through the device (Figure 1) to create a pressure barrier that prevents blood from entering the motor housing. The purge solution becomes systemically bioavailable in the patient after flowing through the device, and the rate of the solution administration is automatically adjusted by the Impella console to target an optimal purge pressure according to device specifications.7 Purge solutions commonly contain heparin, per the manufacturer’s operation recommendations, to reduce the risk of thrombosis, but the optimal anticoagulation strategy when patients with an Impella develop heparin-induced thrombocytopenia (HIT) remains disputed.8

Figure 1.
Figure 1.:
Impella purge solution system (permission obtained from Wiley publishing; Pharmacotherapy 37: 1272–1283, 2017).

The incidence of HIT is highest among patients undergoing cardiac surgeries, with a reported incidence of 1–3%, but patients receiving mechanical circulatory support have an even higher incidence at 10.6%.9,10 Clinicians are likely to encounter scenarios in which alternative anticoagulation strategies are needed to manage HIT whereas a patient is receiving mechanical circulatory support. The use of argatroban purge solutions have been reported in patients with Impellas suspected to have HIT, but to date, no reports describing the use of bivalirudin in these situations have been published.11,12 In these two cases, we describe the use of a bivalirudin-containing purge solution in patients with confirmed HIT who were successfully bridged with Impella support to either durable LVAD or heart transplantation.

Case 1

A 39-year-old man with a medical history of heart failure due to familial cardiomyopathy, automated implantable cardioverter defibrillator (AICD) implantation, atrial fibrillation (AF), and left ventricular thrombus presented to the emergency department with chest pain, weakness, and shortness of breath that had been ongoing for 2 weeks. A pulmonary artery catheter was placed on day 1, which indicated biventricular failure. Dopamine was initiated, and an IABP was placed. Heparin was started empirically because of the patient’s history of left ventricular thrombus. The patient’s platelet count decreased from a baseline of 225,000 cells/microliter on admission to 33,000 cells/microliter on day 5, and because of suspected HIT, an anti-heparin/platelet factor 4 (PF4) antibody test was performed and bivalirudin was initiated. The anti-heparin/PF4 antibody test was positive with an optical density of 2.3 (normal range <0.4) and the serotonin release assay was also positive, confirming the diagnosis of HIT.

The patient’s severely reduced cardiac index despite high-dose inotropes, IABP, and diuretics necessitated the escalation of his mechanical circulatory support device on day 31 to Impella CP, whereas the IABP was left in place. Before the placement of the Impella device, the patient was receiving bivalirudin at a dose of 0.12 mg/kg/h. Based on this, the device infusate solution was formulated to give the patient approximately 50% of the previous dose via the purge solution and allow the systemic solution to be titrated to target the partial thromboplastin time (PTT) according to the institution’s bivalirudin protocol (Table 1). The purge solution was created to contain 100 mg of bivalirudin in 250 ml of 5% dextrose in water (D5W), which was expected to infuse at approximately 17 ml/h based on the current infusate of D5W after the Impella device was placed. The bivalirudin dose from the purge solution alone was determined to be approximately 0.07 mg/kg/h; thus, the systemic bivalirudin was empirically reduced to 0.05 mg/kg/h shortly after changing the device infusate. Of the PTTs drawn, 47 of 63 (74.6%) were within the goal therapeutic range (Figure 2).

Table 1. - Systemic Bivalirudin Protocol
PTT <60 seconds Increase rate by 20%
PTT 60–80 seconds Continue current dose
PTT 81–100 seconds Decrease rate by 25%
PTT 101–120 seconds Decrease rate by 50%
PTT >120 seconds Hold for 1 hour then restart at decreased rate by 50%
PTT, partial thromboplastin time.

Figure 2.
Figure 2.:
Clinical timeline of case 1 (A) and case 2 (B). aPTT, activated partial thromboplastin time; CVVHD, continuous veno-venous hemodialysis; ECMO, extracorporeal membrane oxygenation; IABP, intraaortic balloon pump; LVAD, left ventricular assist device; OHT, orthotopic heart transplantation.

The patient’s platelet count, which had recovered after the cessation of heparin and peaked at 497,000 cells/microliter, declined steadily after the placement of the Impella, likely because of increased platelet destruction. His renal function continued to deteriorate during his hospitalization, and he was placed on continuous veno-venous hemodialysis (CVVHD) on day 37. The initiation of CVVHD increased systemic bivalirudin dose requirements, with a maximum dose of 0.27 mg/kg/h on day 42. On day 39, continued hemodynamic instability led to the placement of peripheral venoarterial extracorporeal membrane oxygenation and removal of the IABP. On day 42, the CVVHD was stopped. On day 48, he was taken to the operating room for the removal of the Impella and off-pump placement of a HeartWare HVAD (Medtronic, Minneapolis, MN) after two sessions of therapeutic plasma exchange to remove anti-heparin/PF4 antibodies. This procedure involved minimal bleeding and no thrombotic complications. His postoperative course was minimally complicated, with his prolonged length of stay because of his extensive rehabilitation required after LVAD implantation. On hospital day 87, he was discharged to an inpatient rehabilitation facility, 39 days after insertion of the durable LVAD.

Case 2

A 69-year-old man with a past medical history of biventricular heart failure because of nonischemic cardiomyopathy, AICD implantation, AF with prior embolic stroke, and chronic kidney disease presented to the emergency department of a community hospital in our health system with worsening dyspnea.

On the day he presented (day 1), a right heart catheterization showed elevated biventricular filling pressures and low cardiac output. An IABP was placed to augment cardiac output and he was placed on CVVHD for his oliguric acute kidney injury. A pulmonary artery catheter placed on day 6 showed improved, but still elevated, biventricular filling pressures and a persistently low cardiac output, so an Impella 5.0 was surgically placed via his right subclavian artery and the IABP was removed. Heparin, delivered to the patient via the purge solution and systemically, was begun after the Impella placement. During this time, the patient’s platelets had gradually decreased from a baseline of 181,000 cells/microliter on admission to 66,000 cells/microliter. An anti-heparin/PF4 antibody test was performed, and it was weakly positive with an optical density of 0.49.

As a result of the positive anti-heparin/PF4 antibody test, the systemic and purge solution heparin were changed to bivalirudin on day 8 to treat for suspected HIT. Because he was on CVVHD at the time of bivalirudin initiation, the patient’s anticipated total dose was 0.03–0.04 mg/kg/h. The purge solution concentration was created to contain 50 mg of bivalirudin in 500 ml of D5W, which was expected to flow at approximately 12 ml/h, based on the current infusate of heparin after the Impella device was placed. The bivalirudin dose from the purge solution alone was determined to be approximately 0.015 mg/kg/h. Systemic bivalirudin was initiated at 0.025 mg/kg/h for a total dose of 0.04 mg/kg/h. Of the PTTs drawn, 28 of 51 (54.9%) were within the goal therapeutic range (Figure 2).

The patient was transferred to our central campus to begin evaluation for advanced therapies on day 9. On day 13, the serotonin release assay resulted positive, confirming the diagnosis of HIT. A repeat anti-heparin/PF4 antibody test on day 15 was positive with an optical density of 3.1. The patient was approved by the medical review board for heart transplantation listing. An offer for transplantation was accepted on day 22, and the patient underwent therapeutic plasma exchange to remove anti-heparin/PF4 antibodies immediately pre transplantation. He underwent successful orthotopic heart transplantation on day 23 with concurrent removal of the Impella device. His course with the Impella 5.0 was uncomplicated, with no thrombotic or major bleeding events noted. His posttransplantation course was uneventful, and on hospital day 36, he was discharged home.


The diagnosis of HIT in patients who undergo cardiovascular surgery or with mechanical circulatory support is challenging because of the potential for multiple causes of thrombocytopenia. These patients, however, have a higher risk of developing HIT than other hospitalized patients, potentially resulting in devastating consequences, and the use of direct thrombin inhibitors in this population poses additional challenges.9,10

Our experience led us to conduct a thorough literature search which yielded one case report and one case series that described the use of argatroban in patients with Impella devices and suspected HIT.11,12 The single case report involved a woman with takotsubo cardiomyopathy who had an Impella 2.5 implanted for treatment of cardiogenic shock. An acute drop in the patient’s platelet count resulted in the initiation of argatroban-containing purge solution, but the results of the anti-heparin/PF4 antibody test and the serotonin release assay both returned negative. The patient was continued on argatroban, and the PTTs were therapeutic from the purge solution alone.11 In the first case of the case series, a woman who underwent coronary artery bypass grafting developed cardiogenic shock because of occlusion of the grafts, necessitating the placement of two Impella 2.5s for biventricular support. An anti-heparin/PF4 antibody test was positive 2 days later, leading to the initiation of systemic argatroban and the removal of heparin from the purge solution; argatroban was added to the purge solution after 48 hours of running only D5W through the device. The second case involved a man who developed cardiogenic shock and, after unsuccessful percutaneous coronary intervention, was placed on an Impella CP with a heparin-containing purge solution. The patient had a positive anti-heparin/PF4 antibody test, and, later, a positive serotonin release assay, so systemic argatroban was initiated. No anticoagulation was used in the purge solution until the flow rate of the purge solution had decreased; out of a concern for device thrombosis, argatroban was added.12

The use of alternative anticoagulant-containing purge solutions in patients requiring percutaneous VADs is not well described in the literature. Our cases differ from those reported in the literature in that only one of the patients described had the diagnosis of HIT confirmed with a serotonin release assay, whereas both of our patients had confirmed HIT. Even more unique is the fact that percutaneous VADs were used as bridges to LVAD implantation and heart transplantation. If the patient’s condition requires use of an alternative anticoagulant, the manufacturer of the Impella recommends systemic use of a nonheparin anticoagulant with a dextrose-only purge solution. Neither argatroban nor bivalirudin is recommended in the purge solution because the Impella has not been tested with these agents.8 The high risk of thromboembolic complications in HIT plus the potentially harmful consequences of thrombosis of the Impella device, however, merit the consideration of adding alternative anticoagulants to the purge solution if the device is expected to remain for an extended time. Our patients received mechanical circulatory support with the Impella device with a bivalirudin purge solution for 17 and 15 days, respectively. Neither patient we described had bleeding or thrombotic complications before their surgeries while receiving bivalirudin, suggesting that bivalirudin can be a safe and effective alternative for Impella anticoagulation if necessary.

The use of split-dose bivalirudin allows for easy titration of the systemically administered bivalirudin to achieve therapeutic PTTs. In the two cases described in this report, our institutional bivalirudin protocol was used for adjusting the systemic bivalirudin (Table 1). The diluent of the purge solution consists of D5W, as the manufacturer no longer recommends higher dextrose concentrations routinely.13 Because of the range of flow rates at which the purge solution can run (typically 7–20 ml/h), a large variation in the dose of anticoagulant delivered to the patient through the purge solution is possible.

In both of our cases, the approach to dosing and initiating the bivalirudin was patient specific. Three variables should be considered when initiating bivalirudin in a patient with a percutaneous VAD: 1) the total dose of bivalirudin to be delivered to the patient; 2) the purge solution flow rate; and 3) the concentration of the bivalirudin-based purge solution. In case 1, the patient was already receiving bivalirudin at a stable dose for confirmed HIT at the time of percutaneous VAD placement. Because both the therapeutic dose of bivalirudin and rate of purge solution were known, an appropriate concentration for supplying approximately half of the bivalirudin dose via the purge solution was calculated. In other circumstances, such as in case 2, the exact dose of bivalirudin that would lead to therapeutic anticoagulation was not known, but the expected dose can be estimated based on calculated creatinine clearance because bivalirudin relies on renal elimination.14,15 The purge solution flow rate was already known, so the bivalirudin-based purge solution was created to deliver approximately half of the anticipated bivalirudin dose via the percutaneous VAD infusate. In both cases, the systemic bivalirudin was initiated at doses intended to deliver the remaining percentage of the known dose, as in case 1, or expected dose, as in case 2.

Because our institution did not have guidelines for the bivalirudin purge solution concentration at the time the aforementioned patients were encountered, two different concentrations were used. The bivalirudin purge solution concentration should be individualized for each patient and can be determined by using the following multistep process. First, the total expected dose of bivalirudin should be established as described previously. The total therapeutic dose can then be decreased by half in anticipation of the patient receiving approximately 50% of the bivalirudin from the purge solution. The purge solution flow rate should be determined based on the current purge solution as displayed on the device console. If the dextrose concentration is unchanged, the purge flow rate should remain close to the same. Finally, create a purge solution with a bivalirudin concentration that supplies approximately 50% of the bivalirudin dose to the patient based on the current purge flow rate. Current batching practices and drug vial sizes at each institution should also be considered when determining the final concentration of the purge solution.


The best management strategy of HIT in patients supported by a percutaneous VAD remains unclear. Our case series is the first to our knowledge that suggests bivalirudin may be a safe and effective anticoagulant when added to purge solutions in patients with Impella percutaneous VAD support who develop HIT. Purge solution concentrations of bivalirudin may be individualized to deliver a portion of the total therapeutic dose required on a patient-specific basis.


The authors would like to thank the staff members of Memorial Hermann – Texas Medical Center and the Center for Advanced Heart Failure for their tireless efforts and exceptional quality of care provided to this challenging patient population. Thanks to Dr. R.M. Sauer Gehring for editorial support.


1. Hochman JS, Boland J, Sleeper LA, et al. Current spectrum of cardiogenic shock and effect of early revascularization on mortality. Results of an International Registry. SHOCK Registry Investigators. Circulation 1995.91: 873–881.
2. van Diepen S, Katz JN, Albert NM, et al.; American Heart Association Council on Clinical Cardiology; Council on Cardiovascular and Stroke Nursing; Council on Quality of Care and Outcomes Research; and Mission: Lifeline: Contemporary management of cardiogenic shock: A scientific statement from the American Heart Association. Circulation 2017.136: e232–e268.
3. Agarwal S, Sud K, Martin JM, Menon V. Trends in the use of mechanical circulatory support devices in patients presenting with ST-segment elevation myocardial infarction. JACC Cardiovasc Interv 2015.8: 1772–1774.
4. Mandawat A, Rao SV. Percutaneous mechanical circulatory support devices in cardiogenic shock. Circ Cardiovasc Interv 2017.10: e004337.
5. Miller PE, Solomon MA, McAreavey D. Advanced percutaneous mechanical circulatory support devices for cardiogenic shock. Crit Care Med 2017.45: 1922–1929.
6. Atkinson TM, Ohman EM, O’Neill WW, et al. A practical approach to mechanical circulatory support in patients undergoing percutaneous coronary intervention. J Am Coll Cardiol Intv 2016;9:871–883.
7. Allender JE, Reed BN, Foster JL, et al. Pharmacologic considerations in the management of patients receiving left ventricular percutaneous mechanical circulatory support. Pharmacotherapy 201737: 1272–1283.
8. Abiomed, Inc: Impella ventricular support systems for use during cardiogenic shock. Impella 2.5, 5.0, LD and Impella CP Instructions for Use and Clinical Reference Manual. 2016.Danvers, MA, Abiomed.
9. Pishko AM, Cuker A. Heparin-induced thrombocytopenia in cardiac surgery patients. Semin Thromb Hemost 2017.43: 691–698.
10. Schenk S, El-Banayosy A, Prohaska W, et al. Heparin-induced thrombocytopenia in patients receiving mechanical circulatory support. J Thorac Cardiovasc Surg 2006.131: 1373–81.e4.
11. Laliberte B, Reed BN. Use of an argatroban-based purge solution in a percutaneous ventricular assist device. Am J Health Syst Pharm 2017.74: e163–e169.
12. Blum EC, Martz CR, Selektor Y, Nemeh H, Smith ZR, To L. Anticoagulation of percutaneous ventricular assist device using argatroban-based purge solution: A case series. J Pharm Pract 2018.31: 514–518.
13. Abiomed, Inc: 5% Dextrose With Heparin As Default Impella Purge Solution. 2015.Danvers, MA, Abiomed.
14. Kiser TH, Fish DN. Evaluation of bivalirudin treatment for heparin-induced thrombocytopenia in critically ill patients with hepatic and/or renal dysfunction. Pharmacotherapy 2006.26: 452–460.
15. Kiser TH, Burch JC, Klem PM, Hassell KL. Safety, efficacy, and dosing requirements of bivalirudin in patients with heparin-induced thrombocytopenia. Pharmacotherapy 2008.28: 1115–1124.

anticoagulation; bivalirudin; cardiogenic shock; percutaneous ventricular assist device; Impella

Copyright © 2020 by the American Society for Artificial Internal Organs