Antithrombotic Therapy for Mechanical Circulatory Support: Time to Throw the Baby (Warfarin) Out with the Bathwater (Aspirin)?

The fully magnetically levitated, centrifugal-flow HeartMate 3 (HM3) left ventricular assist device (LVAD) has revolutionized the landscape of durable mechanical circulatory support. Following HM3 LVAD implant, patients with advanced heart failure can anticipate 5-year survival of nearly 60%.^1,2 Despite improved device technology and outcomes, patients with continuous-flow LVADs have been prescribed a relatively unchanged dual antithrombotic regimen for more than 15 years.

Dual therapy with vitamin K antagonist (VKA) and aspirin is intended to mitigate the risk of thromboembolic events due to activation of the coagulation cascade by the LVAD. However, this regimen burdens patients with a risk of bleeding complications, frequent coagulation monitoring, and dietary changes. These aspects limit quality of life even as improved LVAD technologies continue to extend life. Recently, the Antiplatelet Removal and Hemocompatibility Events With the HeartMate 3 Pump (ARIES-HM3) trial demonstrated that 1) aspirin is not required as part of an antithrombotic regimen that includes warfarin to preserve outcomes, and 2) exclusion of aspirin is associated with a significant decrease in bleeding without increased thromboembolic risk.³ In light of these findings, ongoing research on monotherapy with warfarin or direct oral anticoagulants (DOACs) may provide an avenue to further reduce bleeding risk, burden of care, and lifestyle modifications for LVAD patients.

Per 2023 guidelines, patients with LVADs should receive warfarin with target international normalized ratio (INR) 2.0-3.0 plus aspirin 81-325 mg.⁴ Patients undergo coagulation monitoring as frequently as biweekly when dose-effect relationship is unstable. When INR stabilizes, monitoring typically occurs monthly. Despite careful management, mucosal bleeding in LVAD patients is common and severe bleeding occurs at a rate of ≈25 events/100 patient-years.³ This residual risk and safe withdrawal of aspirin in LVAD patients with bleeding complications motivated the design, rapid enrollment, and successful completion of ARIES-HM3. A common reason for enrollment was patients’ desire to reduce bleeding risk and medication burden.

Despite ARIES findings, the transition of patients to warfarin monotherapy may not address the root of the issue – warfarin, not aspirin, contributes the bulk of bleeding risk and care requirements. Warfarin’s dose-response effect varies greatly between individuals and with dietary changes, and drug-drug interactions are common. Frequent monitoring places a large responsibility on patients and caregivers, especially those who are geographically isolated or have limited transportation. Despite INR monitoring, time in therapeutic range (TTR) is often below 50% in clinical practice,⁵ and below 60% in clinical trials.³

DOACs, namely the factor IIa inhibitor, dabigatran, and factor Xa inhibitors apixaban, edoxaban, and rivaroxaban, have supplanted warfarin in other indications for anticoagulation, including atrial fibrillation⁶ and venous thromboembolism.⁷ Compared to warfarin, DOACs have faster onset and offset, are prescribed in fixed doses, do not require monitoring, have fewer drug and dietary interactions, and are associated with reduced risk of major bleeding.^8,9 Reversal agents are available for the thrombin inhibitor dabigatran (idarucizumab), and the Factor Xa inhibitors apixaban and rivaroxaban (andexanet alfa).

The potential to reduce bleeding complications and burden of care make DOACs a promising alternative to warfarin in LVAD patients. Unfortunately, the negative outcomes of the RE-ALIGN trial of dabigatran versus warfarin in patients with mechanical heart valves⁸ and a pilot trial of dabigatran versus phenprocoumon with HeartWare HVAD¹⁰ raised concern about use of DOACs with LVADs. RE-ALIGN was terminated early due to excess thromboembolic and bleeding events with dabigatran;⁸ the HVAD trial was also terminated due to dabigatran-related thromboembolic events.¹⁰

Despite these negative experiences with dabigatran and older LVAD technology, DOACs that act earlier in the coagulation cascade should still be considered for use with the improved HM3 technology. RE-ALIGN investigators postulated that amplified levels of thrombin during intense contact activation overwhelmed dabigatran, a thrombin inhibitor, while warfarin was effective because it inhibits the contact and common coagulation before significant amplification. Factor Xa inhibitors act earlier in the coagulation cascade and might not have this limitation. Furthermore, HM3 lacks the smooth-to-sintered interface that promoted thrombus formation in HVAD,¹¹ and therefore is associated with lower incidence of pump thrombosis and ischemic stroke.¹² In mechanical heart valves, most thrombi form at the sewing ring prior to endothelization,⁸ but the titanium surface of the HM3 is sintered to promote endothelization.

Apixaban is the most promising factor Xa inhibitor for LVAD anticoagulation because it is associated with lower risks of gastrointestinal bleeding, intracranial hemorrhage, and systemic embolism than rivaroxaban.¹³ Several retrospective studies examined short-term safety and efficacy of apixaban plus aspirin with continuous-flow LVADs. Whitehouse et al.¹⁴ found a similar risk of thrombotic complications and death over 6 months in HM3 patients receiving warfarin or apixaban, and a non-significant reduction in bleeding with apixaban. A small study of apixaban or rivaroxaban in HeartMate II and HVAD patients after warfarin failure found a non-significant reduction in thrombotic and hemorrhagic events/patient-year with DOAC and aspirin.¹⁵ Although promising, these studies were limited by small size, non-randomization, and use of older LVAD technology.

More contemporary studies include the recently completed DOT-HM3 and DOAC LVAD randomized clinical trials (Table). DOT-HM3 randomized 45 patients >3 months post-HM3 implant 2:1 to apixaban 5 mg twice daily or continued therapy with warfarin. At 6 months, there were no thromboembolic events in either arm, and a reduced number of bleeding events with apixaban.¹⁶ DOAC LVAD also showed that anticoagulation with apixaban was feasible in LVAD patients without an excess of hemocompatibility-related adverse events.¹⁷ If pilot studies continue to demonstrate safety, adequately powered trials of apixaban in patients with de novo HM3 and those transitioning from VKA should be conducted to provide conclusive evidence for use of apixaban as first-line anticoagulation in LVAD patients. Furthermore, while apixaban 5 mg twice daily was used in DOT-HM3 and DOAC LVAD, and efficacy of apixaban in AF has not been shown to be associated with blood levels,¹⁸ use of standard dosing with LVAD should include study of trough levels and TTR, and association with thromboembolic and bleeding outcomes.

Table.

Current and Potential Future Studies of Antithrombotic Therapy in LVAD Patients

Current Studies
Study	Design	Population	N	Anticoagulation Therapy	Antiplatelet Therapy	Primary Outcome
DOAC LVAD17 (NCT04865978)	Prospective, randomized controlled, open label	Less than or ≥ 3 months post-HeartMate 3 implant	30	Apixaban 5 mg twice daily vs. warfarin (INR goal 2.0-2.5)	Aspirin 81 mg daily in both groups	Freedom from death or HRAEs (stroke, device thrombosis, bleeding, aortic root thrombus, and arterial non-CNS thromboembolism) at 24 weeks
DOT HeartMate 316 (NCT04974684)	Prospective, single-center, randomized controlled, open label	Minimum 3 months post-HeartMate 3 implant	45	Switch to apixaban vs. continue warfarin	Not specified	Survival free of major HRAEs after 3 months

Possible Future Studies
Design	Study Population	Anticoagulation Therapy	Antiplatelet Therapy	Primary Outcome
Prospective, multi-center, randomized controlled, open label	De novo HeartMate 3 implant	Apixaban 5 mg twice daily vs. warfarin (INR goal 2.0-3.0)	Aspirin 81 mg vs. no aspirin in patients randomized to apixaban	Survival free of major HRAEs
Prospective, multi-center, randomized controlled, open label	Minimum 3 months post-HeartMate 3 implant	Apixaban 5 mg twice daily vs. warfarin (INR goal 2.0-3.0)	Aspirin 81 mg vs. no aspirin in patients randomized to apixaban	Survival free of major HRAEs
Prospective, multi-center, randomized controlled, open label	Minimum 3 months post-HeartMate 3 implant	Switch to novel VKA vs. continue warfarin INR goal 2.0-3.0 for both groups	None	Time in therapeutic range
Prospective, multi-center, randomized controlled, double blind	De novo HeartMate 3 implant	Warfarin vs. novel VKA INR goal 2.0-3.0 for both groups	None	Survival free of major HRAEs

Abbreviations: DOAC, direct oral anticoagulant; DOT, direct oral therapy; HRAEs, hemocompatibility-related adverse events; INR, international normalized ratio; LVAD, left ventricular assist device; VKA, vitamin K antagonist.

One potential limitation of anticoagulation of HM3 patients with apixaban is uncertainty as to what dose of aspirin (if any) should be used. Studies of apixaban with LVADs have mainly examined apixaban/antiplatelet dual therapy, with variable aspirin doses. This question should be addressed with retrospective or prospective observational analyses of warfarin, warfarin/low-dose aspirin, apixaban, apixaban/low-dose aspirin, and apixaban/high-dose aspirin, with HM3. Although apixaban acts earlier in the cascade than dabigatran, and HM3 has improved hemocompatibility compared to older LVADs, another potential limitation is that amplified levels of factor Xa will saturate apixaban. While the positive findings of completed trials of apixaban with HM3 suggest that this is not the case, this risk underscores the importance of ongoing research to develop new DOACs. In the future, optimal anticoagulation for LVAD patients might be achieved by novel DOACs that target the intrinsic pathway¹⁹ or alternative VKAs with metabolism that avoids the cytochrome P450 system.²⁰

Until new agents are developed, however, the factor Xa inhibitor apixaban should be explored for its potential to reduce bleeding risk and care burden for patients with contemporary LVADs. HM3 technology, ARIES, and the DOT-HM3 and DOAC LVAD trials have laid the groundwork for trials of apixaban with HM3 (Table). The LVAD community owes it to their patients to make sure that optimal antithrombotic strategies are developed.