Since their development in the early 1990s, ventricular assist devices (VAD) have fast become a cornerstone in the management of patients with end-stage heart failure.1 The HeartMate II (Thoratec, Pleasanton, CA) and Heartware (Heartware, Framingham, MA) are durable devices used as a bridge to transplantation or as destination therapy in patients who are ineligible for a heart transplant. In the setting of acute cardiogenic shock, nondurable, percutaneous VADs (pVAD), such as the CentriMag (Thoratec, Pleasanton, CA), TandemHeart (CardiacAssist, Pittsburgh, PA), and Impella (Abiomed, Danvers, MA), are used to provide rapid and versatile support as a bridge to recovery or durable support.1–3 Nondurable pVADs are also utilized for support in high-risk percutaneous coronary interventions or electrophysiologic ablation procedures.4–6 These devices allow for the restoration of physiologic flow rates and the stabilization of critically ill patients while treatment decisions are made. Despite their proven benefit in a variety of clinical scenarios, their use is complicated by the need for anticoagulation.1,7,8
Anticoagulation, typically with continuous unfractionated heparin, is titrated on a patient-specific basis because of an intricate balance between the prevention of bleeding and thrombosis. Monitoring of heparin may include the activated clotting time (ACT), activated partial thromboplastin time (aPTT), or plasma heparin assays, such as anti-Xa unfractionated heparin (anti-Xa). At the Medical University of South Carolina (MUSC), anti-Xa values have become the preferred marker over ACT and aPTT for titration of continuous infusion heparin in patients supported by pVADs, as they are influenced to a lesser degree by other patient variables and are shown to correlate best with the true level of heparin anticoagulation.9–15 Most recently, a quality improvement evaluation provided evidence that aPTT values may underestimate the level of anticoagulation in VAD recipients with the authors hypothesizing the exclusivity of anti-Xa levels for VAD recipients.16
Two small studies have demonstrated the feasibility of utilizing heparin protocols for TandemHeart and Impella pVADs to guide anticoagulation, with both protocols relying on aPTT-guided heparin titration.17,18 Currently, there are no published studies specifically evaluating the use of anti-Xa values to guide heparin titration in pVADs. Although MUSC relies on the anti-Xa values to titrate continuous infusion heparin, there was no standardized protocol for management of heparin in pVAD recipients before this evaluation. This lack of standardization can create opportunities for medication errors, morbidity, and mortality. The development of a nursing-driven and pharmacy-driven protocol to be utilized for patients supported by pVADs could potentially reduce errors and improve patient safety. Therefore, a retrospective evaluation of patients supported by the aforementioned devices at MUSC was undertaken with the goal of assessing current pVAD anticoagulation practices and outcomes to potentially identify an optimal anticoagulation strategy.
This was a single-center, retrospective chart review conducted at MUSC. All patients who were at least 18 years of age and who had a CentriMag, TandemHeart, or Impella pVAD implanted between January 1, 2008 and October 31, 2012 were identified and evaluated for inclusion. Patients who received a nondurable pVAD in the cardiac catheterization laboratory for procedural support, received a pVAD with an oxygenator, or were anticoagulated with non-heparin agents were excluded. All methodologies and processes were evaluated and approved by the local Institutional Review Board.
Because the Impella device utilizes a purge solution of heparin in addition to systemic heparinization, patients were divided into two separate cohorts (CentriMag/TandemHeart or Impella), and analyses were conducted within each cohort to assess individual anticoagulation practices to develop heparin titration protocols. At MUSC during the time of this analysis, a goal anti-Xa range of 0.2–0.4 IU/ml, which correlates to an aPTT range of 45–75 seconds based on the institution’s aPTT assay, was considered therapeutic. To further evaluate the utility and correlation of the anti-Xa or aPTT assays in the management of pVADs, the number of values below, above, and within the aforementioned therapeutic goal was analyzed. Within the CentriMag/TandemHeart cohort, patients were subdivided based on whether anti-Xa values or aPTT values were used to titrate heparin in an effort to evaluate potential differences in respective titration practices. However, only patients who had heparin titrated by anti-Xa values were specifically analyzed for respective device anticoagulation protocol development.
Thrombosis and bleeding, as well as levels of anticoagulation before these events, served as markers for the appropriateness of anticoagulation. Thrombosis endpoints included thrombus formation within the device itself or systemic events, such as cardioembolic stroke, thrombotic stroke/transient ischemic attack, myocardial infarction, deep vein thrombosis (DVT), or pulmonary embolism (PE). Major bleeding was defined as an episode of internal or external bleeding that caused death, reoperation, permanent injury, or necessitated transfusion of more than 2 units of red blood cells within a consecutive 24 hour period.2 Both thrombosis and bleeding endpoints were determined through review of progress, surgical, and radiographic reports. Laboratory data and blood product utilization were all reviewed, as well as other anticoagulation testing to assist in further evaluation of thrombosis and bleeding events. Survival endpoints included successful explant after recovery, bridge to transplantation, permanent device implantation, device exchange, or death.
Values for nominal data are reported as percentages, and continuous variables are stated as medians with ranges. Comparison of continuous data was undertaken with nonparametric Mann–Whitney U testing as a means of evaluating the current practice and generating a protocol in addition to hypotheses for further analysis.
During the study period, 24 patients received a CentriMag/TandemHeart pVAD and were considered for inclusion. Seven patients were excluded because of anticoagulation with non-heparin forms of anticoagulation (bivalirudin or argatroban). Baseline demographics for this cohort are presented in Table 1. Five patients received heparin titrated with anti-Xa values, whereas 12 patients had heparin titrations per aPTT values during the course of 13 runs of pVAD support. For the patients included, there was an equal distribution between central and peripheral cannulation with the TandemHeart device being more frequently utilized.
Evaluation of anticoagulation management showed wide disparity between median time to initiation and bolus dose (Table 2), and as expected, there appeared to be a trend toward delayed anticoagulation initiation in patients with central cannulation. Patients who were monitored with anti-Xa values required a higher median dose to maintain therapeutic goals compared with patients monitored with aPTT values. Goal anticoagulation was achieved 68.9% and 43.2% of the time according to anti-Xa and aPTT values, respectively. Patients who were only monitored with aPTT values maintained goal anticoagulation 47.4% of the time.
As shown in Table 2, numerically more patients experienced bleeding and thrombotic events in the cohort monitored with aPTT values than those monitored with anti-Xa values. Survival endpoints were consistent between the two cohorts. In the anti-Xa cohort, bleeding events were only seen in centrally cannulated patients with both having significant chest tube output, and the single thrombotic event occurred in a patient with a previous bleeding episode. As shown in Table 3, median anti-Xa values in the 24 hours preceding major bleeding events were statistically higher than the overall median (p < 0.001). Before the single device thrombosis, the median anti-Xa value was numerically lower than the overall median, although not to a statistically significant degree. The median aPTT value in the anti-Xa cohort preceding the device thrombosis was statistically lower than the overall median, but a significant difference was not seen before bleeding events.
In the aPTT cohort, there was an equal rate of bleeding between centrally (n = 5) and peripherally (n = 4) cannulated patients. Most of the bleeding events in the aPTT group were postoperative bleeding (n = 5) with two patients requiring re-exploration, whereas other bleeding events included femoral cannulation bleeding (n = 2), gastrointestinal (GI) bleeding (n = 1), and oral bleeding (n = 1). Two PEs and one device thrombosis comprised the thrombotic events. For the 24 hours preceding the onset of major bleeding events, the median aPTT value was 49.9 seconds (32.9–150.0 seconds), whereas median aPTT values 24 hours before thrombotic events were 92.2 seconds (75.1–99.6 seconds) for device thrombosis and 65.5 seconds (41.1–101.8 seconds) for systemic thrombosis.
Irrespective of titration strategy, patients who were receiving concomitant aspirin did not experience a higher rate of bleeding.
During the study period, eight patients meeting the inclusion criteria received an Impella device. Anticoagulation was monitored and titrated by anti-Xa values in all eight patients, although per institutional practice, aPTT values were also collected. Baseline demographics for this cohort are presented in Table 4.
Patients with axillary cannulation were noted to have a higher median total heparin rate compared with femoral cannulation (17.3 vs. 15.3 units/kg/hour; p < 0.001; Table 5). Goal anticoagulation was achieved 71.9% of the time per anti-Xa and 52.9% of the time per the additional aPTT monitoring.
Across this cohort, there were no thrombotic events. However, three of the eight patients experienced a combined total of seven major bleeding events (six GI bleeds and one at the cannulation site) with a significantly higher rate occurring in axillary cannulated patients (n = 6) when compared with patients with femoral cannulations despite similar levels of anticoagulation. In the 24 hours preceding these events, the median anti-Xa value was 0.280 IU/ml (0.050–1.380 IU/ml) and was similar to the overall median 0.26 IU/ml (0.050–1.380 IU/ml, p = 0.365). Concomitant aspirin therapy was not associated with a difference in bleeding events.
Heart failure continues to be one of the leading causes of morbidity and mortality in the United States.19 With the advent of nondurable pVADs, patients can be bridged to recovery, a durable device, or transplant potentially improving overall morbidity and mortality. However, there continues to be a lack of consensus on how to optimize anticoagulation in this dynamic population.
In the five patients supported by a CentriMag/TandemHeart device who had heparin titrated based on anti-Xa values, evaluation of median anti-Xa values in the 24 hours before clinical events seemingly provides a goal therapeutic range within which the risk of bleeding and thrombosis may potentially be minimized. However, patients supported with the Impella device had a high rate of bleeding, without any significant thrombosis noted, while maintaining a median anti-Xa value that was previously considered at MUSC to be therapeutic.
With all three devices, this study demonstrated a high percentage of values within the therapeutic range with anti-Xa when compared with aPTT values. Because of the fact that aPTT is more affected by outside influences,9–15 the authors believe that titration based on the anti-Xa assay provides more consistent heparin anticoagulation when compared with the aPTT assay. Nonetheless, there are perceived disadvantages that may limit its overall utilization. Although it has been shown in several reports to have a poor correlation to heparin concentration, the ACT is available as a point-of-care test that can be resulted instantaneously for timely adjustments rather than the lab testing required for the anti-Xa assay.9–11 Also, many clinicians may be less familiar with utilizing the anti-Xa assay over the more commonly used aPTT. With these limitations in mind, it has continued to be the practice at MUSC to additionally obtain aPTT values throughout pVAD support to evaluate for correlation.
Unfortunately, there was a lack of correlation between aPTT and anti-Xa values within the current analysis, which is consistent with previous studies evaluating potential concordance between these two commonly utilized assays.16,20,21 To optimize anticoagulation, evaluation of fibrinogen, antithrombin, and Prothrombin Time/International Normalized Ratio (PT/INR) values can help assess underlying coagulopathies and add a more dynamic value to interpretation. The incorporation of multiple anticoagulation laboratory values to evaluate and recognize hematologic abnormalities has been previously reported.15
The heparin titration protocol that was developed as a result of this research in many ways mirrors the standard of practice at MUSC before this retrospective evaluation. However, a significant change includes the development of new anti-Xa goal therapeutic ranges for all three devices that include both high-intensity and low-intensity goals. Previously, patients supported by each of the three devices initially received heparin titrated to a goal anti-Xa range of 0.2–0.4 IU/ml, but the high level of intrapatient and interpatient variability warrants more individualized therapeutic targets.
The goal anti-Xa range utilized in this analysis differs from commonly reported therapeutic range (0.3–0.7 IU/ml).22–24 Although the ranges developed from this study are significantly lower, the authors believe that the higher level of bleeding events, especially in the Impella cohort, supports this conclusion. Furthermore, the development of high-intensity and low-intensity protocols allows the optimization of goals in patients who are at a higher risk of thrombosis or bleeding.
The protocol developed from the Impella cohort is shown in Appendix I. The high rate of bleeding coupled with the lack of observed thrombosis that was seen in the Impella cohort at anti-Xa values previously considered therapeutic encourages the thought that these patients may require lower anti-Xa values for proper anticoagulation. Thus, the new protocol shifts the initial goal range for patients with an Impella to a lower level of anticoagulation in hopes of limiting bleeding risks, recommending a lower range of 0.15–0.25 IU/ml and a higher range of 0.2–0.3 IU/ml. Because of the high rates of bleeding seen with axillary cannulation, it is recommended that these patients be titrated initially per the lower range protocol.
The protocol developed from the CentriMag/TandemHeart group is shown in Appendix II. Evaluation of the pre- thrombosis and pre-bleeding anti-Xa values in patients supported by the CentriMag/TandemHeart devices seems to suggest that the overall goal range of 0.2–0.4 IU/ml is appropriate as an initial goal. Nonetheless, the new protocol gives clinicians the option for a lower or higher goal range at 0.2–0.3 and 0.3–0.4 IU/ml to provide improved individualization. Additionally, after observing that patients who had central cannulation sites bled more often, it is recommended to initially use the lower intensity goal for these patients.
Although the authors believe that this analysis has provided a better understanding of goal anticoagulation per anti-Xa values for patients on CentriMag/TandemHeart or Impella pVADs, it remains essential that clinicians remain willing to adjust goals for individual patients based on the overall clinical picture. The significant level of heterogeneity that exists in this patient population was profoundly evident during this study. Although patients had a higher percentage of anti-Xa values within the goal range compared with aPTTs, it must be acknowledged that these data do not definitively show that the monitoring of pVAD heparin anticoagulation with heparin anti-Xa values, or any other specific coagulation parameter for that matter, is clearly superior.
This retrospective review is limited by a small sample size, the significant level of heterogeneity that exists in this patient population, and the lack of standardization by which anticoagulation in these individuals was previously titrated. The study was not designed to demonstrate superiority of one particular method but rather to inform protocol development through past experiences. Furthermore, this experience promotes the need to process all available data when making heparin titration decisions in the context of an anti-Xa-based titration protocol.
Although patients receiving non-heparin based products were excluded from this analysis, the utilization of alternative agents including bivalirudin and argatroban merits mentioning. This analysis was employed to develop protocols utilizing anti-Xa values, which are ineffective for monitoring direct thrombin inhibitors (DTIs) because of a difference in mechanism of action. The utilization of DTIs in mechanical circulatory support needs continued exploration as more evidence has established safety with reported benefits but it is beyond the scope of this analysis.25,26
These protocols have been approved by the institution and improved the standardization of care in this patient population. The protocol as a whole should be seen as a starting point for initial dosing and titration followed by continual clinical assessment with refinement of the goals of therapy as necessary. The authors plan to continue to assess for ways to improve this protocol’s performance and validity.
Monitoring of anticoagulation in patients with nondurable pVADs is a complex and intricate endeavor. Anti-Xa assays may facilitate a useful method of titrating heparin anticoagulation in this patient population, but anticoagulation goals should be individualized, and clinicians should remain amenable to altering these goals based on clinical scenarios.
Appendix I: Adult anticoagulation protocol—Impella
Appendix II: Adult anticoagulation protocol—TandemHeart/CentriMag
1. Timms D. A review of clinical ventricular assist devices. Med Eng Phys. 2011;33:1041–1047
2. Miller LW, Pagani FD, Russell SD, et al.HeartMate II Clinical Investigators. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–896
3. Slaughter MS, Rogers JG, Milano CA, et al.HeartMate II Investigators. Advanced heart failure treated with continuous-flow left ventricular assist device
. N Engl J Med. 2009;361:2241–2251
4. Thiele H, Lauer B, Hambrecht R, Boudriot E, Cohen HA, Schuler G. Reversal of cardiogenic shock by percutaneous left atrial-to-femoral arterial bypass assistance. Circulation. 2001;104:2917–2922
5. Kar B, Gregoric ID, Basra SS, Idelchik GM, Loyalka P. The percutaneous ventricular assist device
in severe refractory cardiogenic shock. J Am Coll Cardiol. 2011;57:688–696
6. Sjauw KD, Konorza T, Erbel R, et al. Supported high-risk percutaneous coronary intervention with the Impella 2.5 device the Europella registry. J Am Coll Cardiol. 2009;54:2430–2434
7. Myers TJ. Temporary ventricular assist devices in the intensive care unit as a bridge to decision. AACN Adv Crit Care. 2012;23:55–68
8. Kurien S, Hughes KA. Anticoagulation
and bleeding in patients with ventricular assist devices: walking the tightrope. AACN Adv Crit Care. 2012;23:91–98
9. Smythe MA, Koerber JM, Nowak SN, et al. Correlation between activated clotting time and activated partial thromboplastin time. Ann Pharmacother. 2002;36:7–11
10. Reiner JS, Coyne KS, Lundergan CF, Ross AM. Bedside monitoring of heparin
therapy: comparison of activated clotting time to activated partial thromboplastin time. Cathet Cardiovasc Diagn. 1994;32:49–52
11. Bull BS, Huse WM, Brauer FS, Korpman RA. Heparin
therapy during extracorporeal circulation. II. The use of a dose-response curve to individualize heparin
and protamine dosage. J Thorac Cardiovasc Surg. 1975;69:685–689
12. Vandiver JW, Vondracek TG. Antifactor Xa levels versus activated partial thromboplastin time for monitoring unfractionated heparin
. Pharmacotherapy. 2012;32:546–558
13. Baird CW, Zurakowski D, Robinson B, et al. Anticoagulation
and pediatric extracorporeal membrane oxygenation: impact of activated clotting time and heparin
dose on survival. Ann Thorac Surg. 2007;83:912–919; discussion 919
14. Nankervis CA, Preston TJ, Dysart KC, et al. Assessing heparin
dosing in neonates on venoarterial extracorporeal membrane oxygenation. ASAIO J. 2007;53:111–114
15. Sievert A, Uber W, Laws S, Cochran J. Improvement in long-term ECMO by detailed monitoring of anticoagulation
: a case report. Perfusion. 2011;26:59–64
16. McIlvennan CK, Page RL 2nd, Ambardekar AV, Brieke A, Lindenfeld J. Activated partial thromboplastin time overestimates anti-coagulation in left ventricular assist device
patients. J Heart Lung Transplant. 2014;33:1312–1314
17. Lee Y, Weeks PA. Effectiveness of protocol guided heparin anticoagulation
in patients with the TandemHeart® percutaneous ventricular assist device
. ASAIO J. 2015;61:207–208
18. Jennings DL, Nemerovski CW, Kalus JS. Effective anticoagulation
for a percutaneous ventricular assist device
using a heparin
-based purge solution. Ann Pharmacother. 2013;47:1364–1367
19. Heart Failure Quick Facts. http://www.hfsa.org/heart_failure_facts.asp
. Accessed June 22, 2013
20. Guervil DJ, Rosenberg AF, Winterstein AG, Harris NS, Johns TE, Zumberg MS. Activated partial thromboplastin time versus antifactor Xa heparin
assay in monitoring unfractionated heparin
by continuous intravenous infusion. Ann Pharmacother. 2011;45:861–868
21. Price EA, Jin J, Nguyen HM, Krishnan G, Bowen R, Zehnder JL. Discordant aPTT and anti-Xa values and outcomes in hospitalized patients treated with intravenous unfractionated heparin
. Ann Pharmacother. 2013;47:151–158
22. Marci CD, Prager D. A review of the clinical indications for the plasma heparin
assay. Am J Clin Pathol. 1993;99:546–550
23. Hirsh J, Warkentin TE, Raschke R, et al. Heparin
and low-molecular weight heparin
: mechanisms of action, pharmacokinetics, dosing considerations, monitoring, efficacy, and safety. Chest. 1998;114:489s–510s
24. Olson JD, Arkin CF, Brandt JT, et al. College of American Pathologists Conference XXXI on laboratory monitoring of anticoagulant therapy: laboratory monitoring of unfractionated heparin
therapy. Arch Pathol Lab Med. 1998;122:782–798
25. Sylvia LM, Ordway L, Pham DT, DeNofrio D, Kiernan M. Bivalirudin for treatment of LVAD thrombosis: a case series. ASAIO J. 2014;60:744–747
26. Pappalardo F, Scandroglio AM, Potapov E, et al. Argatroban anticoagulation
induced thrombocytopenia in patients with ventricular assist devices. Minerva Anestesiol. 2012;78:330–335