ABSTRACT
Background
Sodium bicarbonate‐based purge solution (BBPS) is a growing alternative to heparin‐based purge solution (HBPS). However, literature comparing the safety of BBPS to HBPS remains limited. We assessed the safety of HBPS to BBPS in patients with Impella devices by evaluating the rates of major and clinically relevant non‐major bleeding (CRNMB).
Methods
This is a single‐center, retrospective cohort study including patients on Impella CP or 5.5 devices maintained on HBPS or BBPS for ≥ 48 h. The primary outcome was a composite of major bleeding and CRNMB. Secondary endpoints included suspected device thrombosis and systemic complications.
Results
A total of 83 patients met inclusion criteria (48 in the HBPS and 35 in BBPS). There was no difference in the composite incidence of major bleeding and CRNMB between HBPS and BBPS (56.3% vs. 45.7%, p = 0.34) or the incidence of device‐related or systemic complications. However, there was a significantly higher incidence of major bleeding in those with an Impella 5.5 receiving HBPS as compared to BBPS (27.3% vs. 4.8%, p = 0.04).
Conclusion
BBPS may be a safer option in those requiring longer duration of Impella support even in the absence of contraindications to heparin use and offers comparable efficacy to HBPS. The use of BBPS may be a solution towards standardizing and simplifying anticoagulation management in this population. Larger studies are needed to support the durability of Impella devices with BBPS.
Keywords: anticoagulation, cardiogenic shock, cardiology, Impella, mechanical circulatory support, ventricular assist device
Our study of 83 patients with Impella CP or 5.5 devices found that sodium‐bicarbonate based purge solution offers comparable efficacy to heparin‐based purge solution and may be a safer option in those requiring longer duration of Impella support, even in the absence of any contraindications to heparin use.
1. Introduction
Impella CP and Impella 5.5 (Johnson & Johnson MedTech, New Brunswick, NJ) are micro‐axial, continuous‐flow, percutaneous left ventricular assist devices (pLVAD) that are surgically implanted to provide hemodynamic support by pumping blood from the left ventricle into the aorta [1]. These devices are utilized prophylactically in high‐risk percutaneous coronary interventions or as a bridge to recovery in the setting of cardiogenic shock due to various etiologies [1]. Key differences between the Impella CP and 5.5 devices are their site of implantation, extent of hemodynamic support, and duration of use [1, 2]. The Impella CP is typically utilized in situations of acute hemodynamic collapse given its ease of insertion via a retrograde femoral artery approach, provides a maximum average flow of 3.7 L/min and is Food and Drug Administration (FDA) approved for a duration of 4 days [1, 2]. The Impella 5.5 requires axillary cutdown, or direct insertion into the ascending aorta, provides a maximum average flow of 5.5 L/min and is FDA approved for up to 14 days [2]. Aside from these differences, both require a purge solution that runs countercurrent to blood flow to protect the device from biomaterial deposition [1].
The shear stress of the Impella motor risks denaturing and aggregating blood proteins. This can result in biomaterial deposition within the small gaps in the Impella motor purge pathway, which is associated with reduced purge flow and increased purge pressures [3]. If the deposition is not remediated, performance of the device is compromised leading to sub‐optimal ventricular support or device failure. Therefore, a purge solution that flows retrograde to blood flow is required to create a pressure barrier that prevents biomaterial and blood protein deposition within the device (Figure 1) [1, 2]. Historically, a heparin in dextrose 5% in water (D5W) purge solution has been used to prevent purge block. Mechanisms by which heparin‐based purge solution (HBPS) protects the device are thought to be due to heparin's anticoagulant effects, increased viscosity from dextrose which diverts blood away from the Impella motor, and heparin's negative ionic charge which binds to positively charged blood proteins to prevent aggregation [4]. Systemic heparin exposure via the purge solution may be contraindicated in circumstances such as confirmed or suspected heparin induced thrombocytopenia, underlying coagulopathy, or active bleeding. In those with contraindications to heparin, D5W purge solutions have been used. However, D5W alone may be insufficient to prevent blood protein denaturing and purge block during prolonged absences of heparin exposure. Recently, sodium‐bicarbonate based purge solution (BBPS) gained pre‐market approval from the FDA in April 2022 as an alternative to heparin for patients who are intolerant to heparin or have contraindications to heparin [5]. The mechanism of BBPS in preventing purge block is related to its negative charge, which allows for stabilization of blood proteins [4]. BBPS may also exhibit anticoagulant effects by chelating calcium, an essential cofactor in the coagulation cascade, and increasing pH in the motor environment which stabilizes platelet activation [6, 7]. Systemic anticoagulation is still required to prevent thrombus deposition on the Impella catheter, also known as pump thrombosis [8].
FIGURE 1.
The image demonstrates the countercurrent flow of blood and the purge solution, which creates a pressure barrier to protect the Impella motor pathway. [Color figure can be viewed at wileyonlinelibrary.com]
Anticoagulation management in this population remains a challenge. When utilizing HBPS, the heparin exposure via the purge is unpredictable because the flow rate of the purge solution is automatically regulated by the device and fluctuates between 2 to 30 mL/h to maintain purge pressures between 300 and 1100 mmHg [1, 4]. This variability makes systemic heparin titration complicated as the patient is exposed to heparin from two different sources [9]. Despite this challenge, minimal guidance on anticoagulation management in this population exists leading to variable practice across different institutions. In a survey conducted at high‐volume Impella centers in the United States, it was found that 59% of centers adjust systemic heparin based on the heparin exposure via the purge while 41% of centers do not [9]. This unaccounted heparin exposure may increase the risk of bleeding complications in an already at risk population [5, 10]. BBPS has shown to be a growing alternative to HBPS as it allows for simplified anticoagulation management by optimizing systemic anticoagulation without the need to account for heparin exposure via the purge [5]. Case studies have found that BBPS is feasible and maintains pump patency while also reducing the risk of bleeding [11]. Although BBPS is a viable option for management of patients with Impella devices, clinical applications and outcomes remain limited, especially in those that require Impella support for longer than the approved duration. With our study, we aim to assess the safety of HBPS as compared to BBPS in patients with Impella CP or 5.5 devices by evaluating the rates of major bleeding and clinically relevant non‐major bleeding (CRNMB).
2. Methods
2.1. Study Design and Patient Population
This was a single‐center, retrospective cohort study with institutional review board approval that compared the rates of major and CRNMB between HBPS and BBPS in patients with Impella CP or 5.5 pLVAD. Patients were eligible for inclusion if they were on an Impella 5.5 or CP pLVAD from January 2022 to July 2024 and maintained on either HBPS or BBPS for greater than or equal to 48 h. An Impella duration of at least 48 h was required for inclusion to mitigate the influence of early post‐operative or access‐related bleeding, which could confound the assessment of bleeding outcomes related to the purge solution. Patients were excluded from the study if they were initially managed with a direct thrombin inhibitor, received dual mechanical circulatory support (MCS) devices, or were not initiated on systemic heparin within 24 h of BBPS initiation due to ongoing bleeding complications, coagulopathy, or other patient‐specific risk factors precluding anticoagulation initiation.
Patients were included in the BBPS group if they met the following parameters [1]: initiated on BBPS as the index purge solution and maintained on this purge solution for at least 48 h or [2] index purge solution was either HBPS or D5W but transitioned to BBPS within 24 h and maintained on BBPS for at least 48 h. Patients were grouped in the HBPS group if the following criteria were met [1]: initiated on HBPS as the index purge solution and maintained on this purge solution for at least 48 h or [2] index purge solution was either D5W or BBPS but transitioned to HBPS within 24 h and maintained on this purge solution for at least 48 h.
2.2. Outcomes
The primary outcome assessed was a composite incidence of major bleeding and CRNMB according to the International Society on Thrombosis and Hemostasis (ISTH) criteria [12]. Major bleeding was defined as fatal bleeding (suspected or confirmed bleeding as the cause of death), involvement of a critical organ (intracranial, intraspinal, intraocular, retroperitoneal, pericardial, or intramuscular bleeding with compartment syndrome confirmed by imaging such as computed tomography [CT] or echocardiogram), or clinically overt bleeding causing ≥ 2 g/dL decrease in hemoglobin or requiring ≥ 2 units of packed red blood cells within 24 h. CRNMB was any bleed that did not meet major bleeding criteria but resulted in physician guided medical or surgical treatment for bleeding (i.e., administration of reversal agents or required surgical evacuation of bleed) or change in antithrombotic therapy (i.e., interruption for greater than or equal to 2 h or discontinuation of antithrombotic therapy).
Secondary outcomes included the individual components of the primary outcome, device thrombosis, and systemic complications. Device thrombosis was defined as suspected purge block that led to the administration of alteplase via the purge, need for Impella pump replacement, or occurrence of pump pressures ≥ 1000 mmHg and purge flow rates ≤ 3 mL/h requiring medical intervention. These pump pressures and purge flow rate parameters were selected for suspected purge block based on previously published literature [10, 11, 13, 14]. Systemic complications included the incidence of systemic thrombosis and vascular events, such as acute deep vein thrombosis (DVT), acute pulmonary embolism, acute ischemic stroke, and limb ischemia. Diagnosis of systemic embolism was obtained from imaging studies including angiography, CT, or other objective testing. Other outcomes included median activated partial thromboplastin times (aPTT) while on Impella support. All outcomes were evaluated for the duration of Impella support, until purge solution switch (after study arm allocation), switch to systemic anticoagulation with a direct thrombin inhibitor, or initiation of other MCS devices, whichever occurred first.
Subgroup analysis included Impella pressures and purge flows before and after alteplase administration via the purge system and bleeding and thrombotic events between Impella CP and. 5.5 groups. Collection of baseline laboratory parameters and calculation of the sequential organ failure assessment (SOFA) score was conducted within 24 h of Impella placement. The Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) scale, a classification system to categorize severity of advanced heart failure, was also captured for those that had a documented score [15].
2.3. Anticoagulation Management
At our institution, anticoagulation decisions and aPTT goals are assessed daily by a multidisciplinary team involving advanced heart failure physicians, clinical pharmacists, and the intensive care unit team. Anticoagulation is managed by clinical pharmacists following institution‐approved protocols. Available heparin‐based purge concentrations include standard (25 000 units of heparin in 500 mL of D5W), half (12 500 units of heparin in 500 mL of D5W) and quarter (6250 units of heparin in 500 mL of D5W). The purge heparin exposure is accounted for, and systemic heparin is titrated to an aPTT goal of 60–80 s. The available concentration of BBPS is 25 mEq of sodium bicarbonate in 1000 mL of D5W. Systemic anticoagulation with heparin is administered with BBPS and titrated to an aPTT goal of 60–80 s. For suspected purge block or thrombosis, the dose of alteplase administered via the purge system is 4 mg in 50 mL of sterile water (0.08 mg/mL) over 6 h based on published literature [13, 14, 16].
2.4. Statistical Analysis
Patient characteristics are reported as frequencies and percentages for categorical data and as median and interquartile ranges (IQR) for continuous variables. Differences between groups (HBPS vs. BBPS) were determined by Chi‐square or Fisher's exact tests for categorical variables and Wilcoxon rank‐sum test for continuous variables. All data analysis was conducted on Minitab version 22.2. A two‐sided p < 0.05 was considered statistically significant.
3. Results
Between January 2022 and July 2024, 375 patients had an Impella CP or 5.5 placed, of whom 83 met inclusion criteria [48 in the HBPS group and 35 in BBPS] (Figure 2). Predominant reasons for exclusion were Impella placement at an outside facility, device run time for < 48 h or presence of dual MCS devices. Baseline characteristics were well balanced between both groups (Table 1). Median SOFA and INTERMACS scores were similar between the groups. At the time of Impella placement, 11 (22.9%) were on concurrent renal replacement therapy in the HBPS group and 5 (14.3%) in the BBPS group. In terms of initial Impella device, 20 (41.7%) in HBPS and 19 (54.3%) in BBPS received a CP device and the remainder a 5.5 device. Of those that initially had an Impella CP placed, 7 (35%) and 8 (42.1%) were upgraded to an Impella 5.5 device in the HBPS and BBPS groups, respectively. The median duration of Impella device prior to any deviation from study criteria was similar between the HBPS and BBPS groups at 8 (IQR, 4–17) days and 10 (IQR, 4–21) days, p = 0.492. In the BBPS group, the median time to systemic heparin initiation from BBPS start was 4.2 h. The predominant indications for Impella device in both groups were cardiogenic shock and bridge to heart transplant.
FIGURE 2.
The flow diagram represents patient enrollment. Patients with an Impella CP or 5.5 left ventricular device who met inclusion criteria were divided into two groups: heparin‐based or bicarbonate‐based purge solutions. Patients were excluded from the study if they were initially managed with a direct thrombin inhibitor, received dual mechanical circulatory support devices, or were not initiated on systemic heparin within 24 h of bicarbonate‐based purge solution initiation. Abbreviations: BBPS, bicarbonate‐based purge solution; MCS, mechanical circulatory support.
TABLE 1.
Patient and Impella device characteristics.
| HBPS (n = 48) | BBPS (n = 35) | p | |
|---|---|---|---|
| Age (years), median (IQR) | 65 (52–68) | 64 (54–70) | 0.580 |
| Male, n (%) | 38 (79.2) | 32 (91.4) | 0.129 |
| Body mass index (kg/m2), median (IQR) | 29.4 (25.4–33.0) | 28.3 (24.5–31.9) | 0.407 |
| Race, n (%) | |||
| Caucasian | 23 (47.9) | 18 (51.4) | 0.752 |
| Black | 16 (33.3) | 12 (34.3) | 0.928 |
| Asian | 4 (8.3) | 3 (8.6) | 0.969 |
| Other | 5 (10.5) | 2 (5.7) | 0.446 |
| SOFA score, median (IQR) | 8 (6–12) | 9 (8–11) | 0.417 |
| INTERMACS score, median (IQR) a | 2 (2–2) | 2 (2–2) | 0.495 |
| Antiplatelet use, n (%) | |||
| 0 agents | 14 (29.2) | 12 (34.3) | 0.620 |
| 1 agent | 23 (47.9) | 14 (40) | 0.474 |
| 2 agents | 11 (22.9) | 9 (25.7) | 0.769 |
| RRT at Impella placement, n (%) | 11 (22.9) | 5 (14.3) | 0.325 |
| HD | 0 (0) | 1 (2.9) | 0.421 |
| CRRT | 11 (22.9) | 4 (11.4) | 0.179 |
| Laboratory values within 24 h of Impella placement, median (IQR) a | |||
| Hemoglobin (g/dL) | 11.1 (9.2–12.2) | 10.4 (8.7–12.2) | 0.583 |
| Platelets (k/uL) | 152 (113–210) | 190 (124–255) | 0.108 |
| INR | 1.5 (1.2–1.9) | 1.5 (1.3–1.9) | 0.903 |
| Partial thromboplastin time (seconds) | 45.3 (38.9–95.5) | 58.8 (38.2–250) | 0.251 |
| AST (U/L) | 82 (38–313) | 109 (39.5–337) | 0.793 |
| ALT (U/L) | 50 (23–287) | 42 (15–156) | 0.413 |
| Total bilirubin (mg/dL) | 1.1 (0.7–2.5) | 0.9 (0.6–1.5) | 0.227 |
| Initial Impella device, n (%) | |||
| Impella CP | 20 (41.7) | 19 (54.3) | 0.255 |
| Impella 5.5 | 28 (58.3) | 16 (45.7) | 0.255 |
| Upgraded from Impella CP to 5.5, n (%) | 7/20 (35) | 8/19 (42.1) | 0.393 |
| Impella run time (days), median (IQR) | 8 (4–17) | 10 (4–21) | 0.492 |
| Impella run time per device (days), median (IQR) | |||
| Impella CP | 5.1 (2.6–6.8) | 3.5 (2.6–6.2) | 0.407 |
| Impella 5.5 | 11.3 (3.4–20.2) | 15 (7.8–33.3) | 0.291 |
| Impella CP duration > 4 days, n (%) | 12/20 (60) | 7/19 (36.8) | 0.592 |
| Impella 5.5 duration > 14 days, n (%) | 13/35 (37.1) b | 12/24 (50) b | 0.480 |
| Time from BBPS to systemic heparin initiation (hours), median (IQR) | — | 4.2 (1–9.9) | — |
| Indication for Impella, n (%) | |||
| Bridge to OHT | 15 (31.3) | 12 (34.3) | 0.771 |
| Bridge to durable LVAD | 8 (16.7) | 6 (17.2) | 0.954 |
| High‐risk PCI | 5 (10.4) | 4 (11.4) | 0.884 |
| Cardiogenic shock | 20 (41.7) | 13 (37.1) | 0.678 |
Abbreviations: BBPS, bicarbonate based purge solution; CRRT, continuous renal replacement therapy; HD, hemodialysis; IQR, interquartile range; kg, kilogram; LVAD, left ventricular assist device; OHT, orthotopic heart transplant; PCI, percutaneous coronary intervention; RRT, renal replacement therapy; SOFA, Sequential Organ Failure Assessment.
Values only represent values that were available and collected for patients at the time of Impella placement.
Denominator includes those that were upgraded from Impella CP to 5.5.
There was no significant difference in the composite incidence of major bleeding and CRNMB between groups [56.3% in HBPS vs. 45.7% in BBPS, p = 0.343] (Table 2). Numerically, more major bleeding events were noted in the HBPS group as compared to BBPS (11 (22.9%) vs. 4 (11.4%), p = 0.179). There was a statistically significantly higher number of clinically overt bleeding in the HBPS group compared to BBPS (16.7% vs. 2.9%, p = 0.046). Additionally, 1 (2%) patient experienced a fatal bleed due to a procedural complication while on HBPS. The median time from Impella placement to major bleed was 4.6 (IQR, 2.5–6.3) days in HBPS and 0.8 (IQR, 0.6–2.5) days in BBPS. There was no difference in the incidence of CRNMB (37.5% in HBPS vs. 34.3% in BBPS, p = 0.763). Median time from Impella placement to CRNMB was 3.8 (IQR, 1.7–8.3) days in the HBPS group and 3.2 (IQR, 2.6–5.6) days in BBPS. There was no difference in the incidence of suspected purge block/pump thrombosis (12.5% in HBPS vs. 11.4% in BBPS, p = 0.882) or systemic complications (14.6% in HBPS vs. 25.7% in BBPS, p = 0.204). Median time from Impella placement to suspected purge block/pump thrombosis was 14.1 (IQR, 4.3–23.9) days in the HBPS group and 24 (IQR, 16.0–31.7) days in the BBPS group. Median time from Impella placement to systemic thrombosis or limb ischemia was 3.7 (IQR, 2.7–5.3) days in HBPS and 6.6 (IQR, 2.7–33.0) days in BBPS. The average aPTT per patient while included in the study was 65.1 (IQR 56.1–76.3) seconds in HBPS and 69.9 (IQR 63.1–76.7) in BBPS, p = 0.076.
TABLE 2.
Clinical outcomes.
| HBPS (n = 48) | BBPS (n = 35) | p | |
|---|---|---|---|
| Composite incidence, n (%) | 27 (56.3) | 16 (45.7) | 0.343 |
| Major bleeding, n (%) | 11 (22.9) | 4 (11.4) | 0.179 |
| Category of major bleed, n (%) | |||
| Clinically overt bleeding with ≥ 2 g/dL hemoglobin drop or bleeding requiring ≥ 2 unit pRBC | 8 (16.7) | 1 (2.9) | 0.046 |
| Intracranial | 1 (2) | 1 (2.9) | 1.000 |
| Pericardial | 1 (2) | 1 (2.9) | 1.000 |
| Gastrointestinal bleed | — | 1 (2.9) | 0.422 |
| Fatal bleed | 1 (2) | — | 1.000 |
| Time from Impella placement to major bleed (days), median (IQR) | 4.6 (2.5–6.3) | 0.8 (0.6–2.5) | 0.103 |
| PTT at time of major bleed (sec), median (IQR) | 60.5 (52–62.9) | 59.3 (44.5–73.8) | 0.948 |
| Clinically relevant non‐major bleed, n (%) | 18 (37.5) | 12 (34.3) | 0.763 |
| Bleeding requiring physician guided treatment, n (%) | 2 (4.2) | 2 (5.7) | 0.745 |
| Medical | 2 (4.2) | 1 (2.9) | |
| Surgical | — | 1 (2.9) | |
| Change in anticoagulant therapy, n (%) | 16 (33.3) | 10 (28.6) | 0.644 |
| Anticoagulation held | 13 (27.1) | 9 (25.7) | |
| Dose decreased | — | 1 (2.9) | |
| Heparin purge switched to BBPS | 2 (4.2) | — | |
| aPTT goal reduced | 1 (2.1) | — | |
| Time from Impella placement to CRNMB (days), median (IQR) | 3.8 (1.7–8.3) | 3.2 (2.6–5.6) | 0.816 |
| PTT at time of CRNMB (sec), median (IQR) | 69.5 (61.4–75.6) | 67.3 (50.8–71.9) | 0.278 |
| Suspected purge block/thrombosis, n (%) | 6 (12.5) | 4 (11.4) | 0.882 |
| Alteplase administered via purge, n (%) | 5 (10.4) | 4 (11.4) | 0.884 |
| Impella exchange with confirmed clot, n (%) | 2 (4.2)* | 1 (2.9)* | 0.752 |
| Pump pressures ≥ 1000 mmHg and purge flow rates ≤ 3 mL/h requiring intervention, n (%) | 1 (2.1) | — | — |
| Time from Impella placement to purge block/thrombosis (days), median (IQR) | 14.1 (4.3–23.9) | 24 (16–31.7) | 0.749 |
| PTT at time of suspected purge block/thrombosis (sec), median (IQR) | 46 (39.5–63.3) | 68.3 (61.1–76.8) | 0.241 |
| LDH at suspected Impella thrombosis (U/L), median (IQR) | 433 (284–1462) | 905 (639–975) | 1.000 |
| Systemic complications, n (%) | 7 (14.6) | 9 (25.7) | 0.204 |
| Systemic Thrombosis, n (%) | |||
| Acute deep vein thrombosis | 3 (6.3) | 1 (2.9) | 0.476 |
| Acute ischemic stroke | 1 (2.1) | 2 (5.7) | 0.745 |
| Limb ischemia, n (%) | 3 (6.3) | 6 (17.1) | 0.115 |
| Time from Impella placement to systemic complication (days), median (IQR) | 3.7 (2.7–5.3) | 6.6 (2.7–33) | 0.525 |
| PTT at time of systemic complication (sec), median (IQR) | 55.6 (50.4–57.7) | 63.4 (38.1–75.2) | 0.444 |
| Average PTT per patient, median (IQR) | 65.1 (56.1–76.3) | 69.9 (63.1–76.7) | 0.076 |
| PTT collected per patient per day, median (IQR) | 3.3 (2.8–3.6) | 2.8 (2.4–3.2) | 0.014 |
Abbreviations: LDH, lactate dehydrogenase; PTT, partial thromboplastin time.
Alteplase via Impella purge system also administered.
A total of 9 patients received alteplase via the Impella purge system for suspected purge block (5 in HBPS and 4 in BBPS). Median purge pressures and flows prior to alteplase administration were 863 (IQR, 567–917) mmHg and 3 (IQR, 2–5) mL/h, respectively (Table 3). Median lactate dehydrogenase prior to alteplase administration was 905 (IQR, 403–1253) U/L. Median purge pressures and flows after alteplase administration improved to 552 (IQR, 462–594) mmHg and 9 (IQR, 7–13) mL/h, respectively.
TABLE 3.
Purge pressures and flows before and after alteplase administration.
| Alteplase administered via purge system (n = 9) a | ||
|---|---|---|
| Pre‐tPA | Post‐tPA | |
| Purge pressure (mmHg), median (IQR) | 863 (567–917) | 552 (462–594) |
| Purge flow (mL/h), median (IQR) | 3 (2–5) | 9 (7–13) |
| LDH (U/L), median (IQR) | 905 (403–1253) | — |
Abbreviations: LDH, lactate dehydrogenase; tPA, alteplase.
Suspected purge block/pump thrombosis requiring administration of alteplase 4 mg/50 mL (0.08 mg/mL) in sterile water via Impella purge.
In the Impella CP group, there was no difference in the incidence of major bleed, CRNMB, systemic thrombosis, limb ischemia or suspected purge block/pump thrombosis between HBPS vs. BBPS (Table 4). The median duration on Impella CP was 5.2 (IQR, 3.2–6.8) days and 3.6 (IQR, 2.7–6.3) days in the HBPS and BBPS groups, respectively. In the Impella 5.5 group, major bleeding was significantly higher in those that received HBPS as compared to BBPS (27.3% vs. 4.8%, p = 0.038). In addition, there was no difference seen with CRNMB, systemic thrombosis, limb ischemia, or suspected purge block/pump thrombosis between groups in those with an Impella 5.5. The median duration on Impella 5.5 was 11.7 (IQR, 3.4–29.6) days in HBPS and 16 (IQR, 9.1–32.5) days in BBPS.
TABLE 4.
Clinical outcome stratification by device and purge solution.
| Impella CP | Impella 5.5 | |||||
|---|---|---|---|---|---|---|
| HBPS (n = 15) | BBPS (n = 14) | p | HBPS (n = 33) | BBPS (n = 21) | p | |
| Major bleed, n (%) | 2 (13.3) | 3 (21.4) | 0.564 | 9 (27.3) | 1 (4.8) | 0.038 |
| Clinically overt bleeding with ≥ 2 g/dL hemoglobin drop or bleeding requiring ≥ 2 unit pRBC | 2 (13.3) | 1 (7.1) | 0.584 | 6 (18.2) | 0 (0) | 0.038 |
| Intracranial | 0 (0) | 1 (7.1) | 0.483 | 1 (3) | 0 (0) | 1.000 |
| Pericardial | — | — | 1 (3) | 1 (4.8) | 1.000 | |
| Gastrointestinal bleed | 0 (0) | 1 (7.1) | 0.483 | — | — | |
| Fatal bleed | — | — | 1 (3) | 0 (0) | 1.000 | |
| Clinically relevant non‐major bleed, n (%) | 4 (26.7) | 5 (35.7) | 0.599 | 14 (42.4) | 7 (33.3) | 0.504 |
| Bleeding requiring physician guided treatment | — | — | 2 (6.1) | 2 (9.5) | 0.636 | |
| Change in anticoagulant therapy | 4 (26.7) | 5 (35.7) | 0.599 | 12 (36.4) | 5 (23.8) | 0.333 |
| Systemic complication, n (%) | 2 (13.3) | 3 (21.4) | 0.564 | 5 (15.2) | 6 (28.6) | 0.233 |
| Limb ischemia | 1 (6.7) | 3 (21.4) | 0.249 | 2 (6.1) | 3 (14.3) | 0.309 |
| Acute deep vein thrombosis | 1 (6.7) | 0 (0) | 1.000 | 2 (6.1) | 1 (4.8) | 0.839 |
| Acute ischemic stroke | — | — | 1 (3) | 2 (9.5) | 0.310 | |
| Suspected purge block/thrombosis, n (%) | 0 (0) | 1 (7.1) | 0.483 | 6 (18.2) | 3 (14.3) | 0.708 |
| Alteplase administered via purge | 0 (0) | 1 (7.1) | 0.483 | 5 (15.2) | 3 (14.3) | 0.930 |
| Impella exchange with confirmed clot | — | 2 (6.1)* | 1 (4.8)* | 0.839 | ||
| Pump pressures ≥ 1000 mmHg and purge flow rates ≤ 3 mL/h requiring intervention | — | 1 (3) | 0 (0) | 1.000 | ||
| Duration of device (days), median (IQR) | 5.2 (3.2–6.8) | 3.6 (2.7–6.3) | 0.471 | 11.7 (3.4–29.6) | 16 (9.1–32.5) | 0.214 |
Alteplase via Impella purge system was also administered.
4. Discussion
Limited studies have been published comparing the safety of HBPS to BBPS for Impella pLVADs. Our study demonstrated no difference in the composite incidence of major bleeding and CRNMB between the two purge solutions. Secondary analysis also found no difference in device or systemic thrombosis and vascular complications. However, our subgroup analysis comparing clinical outcomes between Impella CP and 5.5 devices demonstrated a significantly higher incidence of major bleeding in those that had an Impella 5.5 device and received HBPS, highlighting that BBPS may be a safer option in those requiring a longer duration of Impella support even in the absence of any contraindications to heparin use.
Several complications can be associated with Impella devices, such as bleeding, stroke, limb ischemia, purge block, and pump thrombosis [17, 18]. Bleeding complications are multifactorial, including the use of anticoagulation to prevent thromboembolism and increased proteolysis of von Willebrand factor from mechanical shear stress, both of which result in platelet dysfunction [17]. Although there was no significant difference in the composite incidence of major bleeding and CRNMB between groups, a higher number of major bleeds were noted in those that received HBPS. Additionally, a significant subset of clinically overt major bleeds occurred in the HBPS group. The incidence of major bleed was lower in our study than reported in the literature. A retrospective study that included patients on various Impella devices for at least 24 h found an incidence of major bleeding of 65.3% in HBPS vs. 27.9% in BBPS (p < 0.05) [10] However, differences in this study compared to our study design are that all Impella devices (i.e., 2.5, CP, 5.5, RP, and bi‐ventricular support) were included and systemic anticoagulation with heparin was not required in the BBPS arm. These differences could have contributed to the higher incidence of major bleeding seen between the study groups compared to ours. A retrospective descriptive study assessing the prevalence of bleeding as defined by the Bleeding Academic Research Consortium (BARC) criteria in patients with Impella devices receiving HBPS in combination with systemic heparin for at least 12 h found a bleeding event rate of 63%, of which BARC type 3a cases comprised most of the events at 26% [19]. Predominantly, 71% of the patients had an Impella CP in this study. Despite differences in study designs and definitions of major bleeding, a trend towards higher bleeding events is noted in our study and existing literature with the use of HBPS as compared to BBPS.
Since the FDA approved BBPS in April 2022, our institution has standardized the use of BBPS for all patients with Impella devices, allowing for anticoagulation to be provided through a single source. This alleviates the need to account for heparin via the purge system, which has variable exposure, and simplifies overall management of anticoagulation in this patient population. A retrospective cohort study found a higher rate of supratherapeutic aPTTs with the use of HBPS and systemic heparin as compared to BBPS (29% vs. 19%, p = 0.62) [10]. A study describing outcomes with HBPS in combination with systemic heparin with a target aPTT goal of 55–75 s found the percentage of aPTT values within target range per patient of 44.5% [19]. Given the potential for fluctuations in aPTTs with two sources of heparin, BBPS may be a solution towards standardizing and simplifying anticoagulation management in this population.
Systemic complications such as acute venous thromboembolism, acute ischemic stroke, and limb ischemia are other consequences associated with Impella devices [18]. Venous thromboembolism and ischemic strokes can be caused by suboptimal anticoagulation, thrombus formation and dislodgement, and prolonged duration of Impella support [20, 21]. Vascular complications such as limb ischemia have been reported to occur at an incidence of 13.5% and are associated with device positioning precluding blood flow to the extremities, thrombus formation leading to distal embolization, and patient‐related risk factors such as vessel size and peripheral vascular disease [22, 23]. Our study found no difference in the incidence of acute DVT, acute ischemic stroke, or limb ischemia between the HBPS and BBPS groups. There is limited published literature comparing the incidence of stroke and vascular events with HBPS and BBPS, but a retrospective study including 179 patients found no difference in the incidence of stroke and vascular events (17.9% HBPS vs. 33.3% BBPS, p = 0.254) [24]. Our study adds to the growing body of literature, showcasing the efficacy and safety of BBPS as compared to HBPS, and supporting its use as standard of care.
Since its recent FDA approval, the longevity of the Impella device with the use of BBPS is not clearly reported. In our study, we found that the time from Impella placement to suspected purge block/pump thrombosis was 14.1 days in HBPS and 24 days in BBPS. In clinical practice, the Impella devices are often utilized for longer periods of time beyond package insert recommendations as a bridge to heart transplant or durable LVAD [11, 25]. The smaller diameter of the Impella CP results in increased rotational speed in order to maintain a set linear speed, which causes higher shear stress at the impellar blade; thereby, increasing the risk of hemolysis and acute kidney injury with Impella CP [26]. Studies have also found increased hemorrhagic complications and stroke with Impella CP as compared to 5.5 [27]. Our subgroup analysis comparing clinical outcomes with HBPS and BBPS in Impella CP found no difference in the incidence of major bleed, CRNMB, systemic complications or suspected purge block/pump thrombosis. However, a significantly higher incidence in major bleeding was seen in those that were on Impella 5.5 and received HBPS, primarily driven by overt clinical bleeding (defined by a hemoglobin drop of ≥ 2 g/dL or a requirement for ≥ 2 units of pRBCs). Five out of six of these major bleeding events were related to Impella access‐site bleeding or hematoma at the Impella insertion site. Notably, all of these events occurred in the HBPS cohort with a median time of 3 days from Impella 5.5 placement. This suggests that during the peri‐procedural period, exposure to heparin via the purge solution may contribute to bleeding. Conversely, BBPS can help minimize bleeding risk as it allows for more intentional management of systemic anticoagulation, which can be temporarily held during the peri‐procedural period and resumed once bleeding risk subsides or hemostasis is achieved. Additionally, the median duration on Impella CP ranged from 4 to 5 days, whereas median duration on Impella 5.5 ranged from 12 to 16 days. Therefore, for prolonged duration of Impella devices, BBPS may be a safer option to minimize the risk of major bleeding.
There were several limitations to note within our study. Time within target aPTT was not captured as certain patients may require aPTT goal adjustments based on individual bleeding and thromboembolic risks. Furthermore, anticoagulation was primarily monitored via aPTT per our institution protocol without comparison to alternative parameters such as anti‐Xa. While we recognize the potential for discordance between aPTT and anti‐Xa levels in critically ill and MCS populations, the clinical implications of this discrepancy require further research. Other limitations include the patient's pre‐existing bleed risk based on comorbid conditions and variable transfusion thresholds which could have impacted our primary outcome. Additionally, the decision between utilizing HPBS and BBPS prior to institution standardization of BBPS use was left to the discretion of the provider; therefore, we could not control for selection bias introduced due to variations in clinical practices. Lastly, given the retrospective design of the study, we were limited to chart review and note documentation to capture device and patient outcomes.
5. Conclusion
This study compared the safety of HBPS to BBPS for Impella CP and 5.5 devices and demonstrated no difference in the incidence of major bleeding and clinically relevant non‐major bleeding between the two purge solutions. Secondary analysis also found no difference in device or systemic thrombosis and vascular complications. However, our subgroup analysis found a significantly higher incidence of major bleeding in those that had an Impella 5.5 device and received HBPS. The findings of our study demonstrate that BBPS may be a safer option in those requiring longer duration of Impella support even in the absence of any contraindications to heparin use and offers comparable efficacy to HBPS. The use of BBPS may be a solution towards standardizing and simplifying anticoagulation management in this population. However, larger studies are needed to further support the longevity of Impella devices with BBPS and its impact on patient outcomes.
Author Contributions
Data collection, data analysis/interpretation, statistics, drafting article: Priya J. Patel. Data interpretation, critical revision of article, approval of article: Jesse E. Harris. Critical revision of article, approval of article: Kevin R. Donahue. Data interpretation, approval of article. Ju H. Kim. Data interpretation, critical revision of article, approval of article: Mariah I. Sigala.
Funding
The authors have nothing to report.
Ethics Statement
The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the local institutional review board at Houston Methodist Research Institute (no. PRO00038255) on September 4, 2024, with the need for written informed consent waived.
Conflicts of Interest
The authors declare no conflicts of interest.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Johnson & Johnson MedTech , “Impella Heart Pumps,” (2025), https://www.heartrecovery.com/en‐us/products‐and‐services/impella.
- 2. Zein R., Patel C., Mercado‐Alamo A., Schreiber T., and Kaki A., “A Review of the Impella Devices,” Interventional Cardiology 17 (2022): e05, 10.15420/icr.2021.11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Johnson & Johnson MedTech , “Heart Recovery ‐ Education Library,” (2025), https://www.heartrecovery.com/en‐us/education/education‐library.
- 4. Beavers C. J., DiDomenico R. J., Dunn S. P., et al., “Optimizing Anticoagulation for Patients Receiving Impella Support,” Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy 41 (2021): 932–942, 10.1002/phar.2629. [DOI] [PubMed] [Google Scholar]
- 5. Johnson & Johnson MedTech , “Heparin‐Free Purge Solution,” (2025), https://www.heartrecovery.com/en‐us/education/heparin‐free‐purge.
- 6. El‐Hennawy A. S., Frolova E., and Romney W. A., “Sodium Bicarbonate Catheter Lock Solution Reduces Hemodialysis Catheter Loss due to Catheter‐Related Thrombosis and Blood Stream Infection: An Open‐Label Clinical Trial,” Nephrology, Dialysis, Transplantation 34 (2019): 1739–1745, 10.1093/ndt/gfy388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ammann K. R., Outridge C. E., Roka‐Moiia Y., et al., “Sodium Bicarbonate as a Local Adjunctive Agent for Limiting Platelet Activation, Aggregation, and Adhesion Within Cardiovascular Therapeutic Devices,” Journal of Thrombosis and Thrombolysis 56 (2023): 398–410, 10.1007/s11239-023-02852-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Moretz J. D., Corbett S., and Bilazarian S., “Comment: Analysis of Bicarbonate‐Based Purge Solution in Patients With Cardiogenic Shock Supported via Impella Ventricular Device,” Annals of Pharmacotherapy 57 (2023): 1120–1121, 10.1177/10600280221139458. [DOI] [PubMed] [Google Scholar]
- 9. Reed B. N., DiDomenico R. J., Allender J. E., et al., “Survey of Anticoagulation Practices With the Impella Percutaneous Ventricular Assist Device at High‐Volume Centers,” Journal of Interventional Cardiology 2019 (2019): 1–6, 10.1155/2019/3791307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Bergen K., Sridhara S., Cavarocchi N., Silvestry S., and Ventura D., “Analysis of Bicarbonate‐Based Purge Solution in Patients With Cardiogenic Shock Supported via Impella Ventricular Assist Device,” Annals of Pharmacotherapy 57 (2023): 646–652, 10.1177/10600280221124156. [DOI] [PubMed] [Google Scholar]
- 11. Sigala M. I., Harris J. E., Morton C., Donahue K. R., and Kim J. H., “A Case Series Analysis of Bicarbonate‐Based Purge Solution Administration via Impella Ventricular Assist Device,” American Journal of Health‐System Pharmacy 81 (2024): e115–e121, 10.1093/ajhp/zxad278. [DOI] [PubMed] [Google Scholar]
- 12. Schulman S., Kearon C., and Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis , “Definition of Major Bleeding in Clinical Investigations of Antihemostatic Medicinal Products in Non‐Surgical Patients,” Journal of Thrombosis and Haemostasis 3 (2005): 692–694, 10.1111/j.1538-7836.2005.01204.x. [DOI] [PubMed] [Google Scholar]
- 13. Succar L., Donahue K. R., Varnado S., and Kim J. H., “Use of Tissue Plasminogen Activator Alteplase for Suspected Impella Thrombosis,” Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy 40 (2020): 169–173, 10.1002/phar.2356. [DOI] [PubMed] [Google Scholar]
- 14. Tunney R. K., Tidwell W. P., Kumar A. A., and Brinkley M., “Use of Tissue Plasminogen Activator Alteplase Purge Solution in an Impella 5.5 Heart Pump,” ASAIO Journal 67 (2021): e163–e165, 10.1097/MAT.0000000000001349. [DOI] [PubMed] [Google Scholar]
- 15. Alba A. C., Rao V., Ivanov J., Ross H. J., and Delgado D. H., “Usefulness of the INTERMACS Scale to Predict Outcomes After Mechanical Assist Device Implantation,” Journal of Heart and Lung Transplantation 28 (2009): 827–833, 10.1016/J.HEALUN.2009.04.033. [DOI] [PubMed] [Google Scholar]
- 16. Am B., Copeland H., Deswal A., et al., “The International Society for Heart and Lung Transplantation/Heart Failure Society of America Guideline on Acute Mechanical Circulatory Support,” Journal of Cardiac Failure 29 (2023): 304–374, 10.1016/J.CARDFAIL.2022.11.003. [DOI] [PubMed] [Google Scholar]
- 17. Saito S., Okubo S., Matsuoka T., et al., “Impella – Current Issues and Future Expectations for the Percutaneous, Microaxial Flow Left Ventricular Assist Device,” Journal of Cardiology 83 (2024): 228–235, 10.1016/j.jjcc.2023.10.008. [DOI] [PubMed] [Google Scholar]
- 18. Ali J. M. and Abu‐Omar Y., “Complications Associated With Mechanical Circulatory Support,” Annals of Translational Medicine 8 (2020): 835, 10.21037/atm.2020.03.152. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Jiang C., Stuart M., Makowski C., Jennings D. L., and To L., “Safety and Efficacy of a Percutaneously Inserted Ventricular Support Device Purge Solution Heparin 25 U/mL,” Annals of Pharmacotherapy 55 (2021): 174–180, 10.1177/1060028020944831. [DOI] [PubMed] [Google Scholar]
- 20. Novoa I., Hernandez G., Parmar R., Breton J. N., Blumer V., and Chaparro S., “Incidence of Stroke Increases With Prolonged Use of Impella Percutaneous Left Ventricular Assist Device,” Journal of the American College of Cardiology 71 (2018): A808, 10.1016/S0735-1097(18)31349-4. [DOI] [Google Scholar]
- 21. Hassett C. E., Cho S.‐M., Hasan S., et al., “Ischemic Stroke and Intracranial Hemorrhages During Impella Cardiac Support,” ASAIO Journal 66 (2020): e105–e109, 10.1097/MAT.0000000000001132. [DOI] [PubMed] [Google Scholar]
- 22. Abaunza M., Kabbani L. S., Nypaver T., et al., “Incidence and Prognosis of Vascular Complications After Percutaneous Placement of Left Ventricular Assist Device,” Journal of Vascular Surgery 62 (2015): 417–423, 10.1016/j.jvs.2015.03.040. [DOI] [PubMed] [Google Scholar]
- 23. Sinning J.‐M., Ibrahim K., Schröder J., Sef D., and Burzotta F., “Optimal Bail‐Out and Complication Management Strategies in Protected High‐Risk Percutaneous Coronary Intervention With the Impella,” European Heart Journal Supplements: Journal of the European Society of Cardiology 24 (2022): J37–J42, 10.1093/eurheartjsupp/suac064. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Betts B., Moser M., Hee Kim J., et al., “949: Comparison of Bicarbonate‐Based Versus Heparin‐Based Impella Purge Solution,” Critical Care Medicine 53 (2025): e949, 10.1097/01.ccm.0001102460.74433.c2. [DOI] [Google Scholar]
- 25. Ramzy D., Soltesz E., and Anderson M., “New Surgical Circulatory Support System Outcomes,” ASAIO Journal 66 (2020): 746–752, 10.1097/MAT.0000000000001194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Van Edom C. J., Gramegna M., Baldetti L., et al., “Management of Bleeding and Hemolysis During Percutaneous Microaxial Flow Pump Support,” JACC. Cardiovascular Interventions 16 (2023): 1707–1720, 10.1016/j.jcin.2023.05.043. [DOI] [PubMed] [Google Scholar]
- 27. Philipson D. J., Cohen D. J., Fonarow G. C., and Ziaeian B., “Analysis of Adverse Events Related to Impella Usage (From the Manufacturer and User Facility Device Experience and National Inpatient Sample Databases),” American Journal of Cardiology 140 (2021): 91–94, 10.1016/j.amjcard.2020.10.056. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.