Anticoagulation in the HeartMate 3 Left Ventricular Assist Device: Are We Finally Moving the Needle?

Left ventricular assist device (LVAD) therapy significantly improve survival and quality of life of patients with medical therapy refractory advanced heart failure, yet remains burdened by an undesirably high rate of complications.1 Despite a clear reduction in pump thrombosis exhibited in the HeartMate 3 (HM3) LVAD, hemocompatibility-related adverse events, especially strokes and complications of chronic anticoagulation, remain hurdles toward optimal outcomes.2 Standard anticoagulation with a vitamin-K antagonist, most commonly warfarin, has long been problematic because of the need of frequent International Normalized Ratio monitoring, dietary restrictions, and costly hospitalizations and complications related to sub- and supra-therapeutic International Normalized Ratio levels. Newer direct-acting oral anticoagulants (DOACs) are either direct thrombin inhibitor (dabigatran) or direct factor Xa inhibitors (rivaroxaban, apixaban, edoxaban, and betrixaban). DOACs have been proven safe and effective for indications such as nonvalvular atrial fibrillation and venous thromboembolism and have the advantage of obviating the need for therapeutic level monitoring.3 Potential downsides of DOAC therapy include significantly higher drug cost, and that missing a single medication dose results in excess thrombotic risk, in light of a much shorter half-life compared with warfarin. Despite these issues, it is not surprising that DOACs have become an attractive alternative anticoagulation option to study in an LVAD population. However, the initial enthusiasm for DOAC use in LVAD patients was dampened after disappointing results of two randomized clinical trials (RCTs) involving implanted intracardiac devices. The RE-ALIGN (Randomized, Phase II Study to Evaluate the Safety and Pharmacokinetics of Oral Dabigatran Etexilate in Patients after Heart Valve Replacement) trial evaluated warfarin versus dabigatran anticoagulation in patients with mechanical heart valves and was terminated prematurely because of an excess of thromboembolic and bleeding events in the dabigatran group.4 The only randomized control trial of DOAC in LVAD patients, which was designed before the RE-ALIGN results, compared phenprocoumon versus dabigatran in HeartWare ventricular assist device (HVAD) patients.5 This trial was also prematurely terminated after enrollment of 16 patients due to an excess of thromboembolic events in the dabigatran group. In this issue of the ASAIO Journal, Whitehouse et al.6 presents the results of a retrospective, non-randomized analysis of HM3 patients who received either warfarin or apixaban for anticoagulation. Patients at a single academic center were either switched from warfarin to apixaban because of compliance issues or based on shared decision-making after apixaban was offered by the treating physicians as an alternative to warfarin to all new implants starting in July 2020. A total of 35 patients (20 in the warfarin group and 15 in the apixaban group) were included, and the 2 groups were similar at baseline. At 6 months after implant, there was no statistically significant difference in death, stroke, bleeding, or other thrombotic complications between the warfarin and apixaban groups. How do we interpret the results of this study in the historical context of excess thromboembolism in the DOAC group in previous RCTs of LVAD patients and patients with mechanical heart valves? With caution and some optimism. It must be emphasized that the current study is a small, single-center, retrospective, non-randomized cohort, and thus should be considered exploratory and hypothesis-generating. The follow-up time was only 6 months and most outcomes happened in two or fewer patients, severely limiting the statistical power to detect a potential difference between the warfarin and apixaban groups. There was also no data on time in therapeutic range in the warfarin group, making it difficult to know what quality of warfarin therapy apixaban was compared with. Taking a step back, thrombotic complications of LVADs are closely linked to activation of the contact coagulation pathway by the metallic surface of the LVAD, which is absent in the case of atrial fibrillation. It is possible that intracardiac metallic devices trigger inherently different coagulation cascades that explain the different outcomes of DOACs in LVAD and mechanical heart valve trials versus atrial fibrillation trials. Nevertheless, there is much reason for optimism regarding the potential of apixaban in HM3 patients. First of all, the only RCT of DOAC in LVAD patients published by Andreas et al.5 studied dabigatran in HVAD patients. There is now evidence that for stroke prevention in atrial fibrillation, apixaban may have the lowest stroke and bleeding rates compared with other DOACs.7,8 So it is possible that apixaban may also have lower stroke and bleeding rates than dabigatran in LVAD patients. Second, in the RCT by Andreas et al., all four patients in the dabigatran group who experienced thromboembolic events had device thrombosis, three while on dabigatran, and one discovered shortly after switch to phenprocoumon. Given that HM3 has shown a lower rate of pump thrombosis, the excess thromboembolic rates of DOAC over warfarin may be blunted in patients with a HM3. Last, in a recent ex vivo study of HVAD pump thrombosis using multiple anticoagulation agents, apixaban had similar thrombosis times compared with warfarin whereas dabigatran had thrombosis times similar to controls without anticoagulation.9 This study further bolsters the case for safety of apixaban anticoagulation in LVAD patients. The safety and efficacy of wide-spread use of DOAC anticoagulation in LVAD patients needs to be definitely evaluated by an adequately powered prospective RCT. Future studies of DOAC versus warfarin in LVAD patients should also need to assess the impact of warfarin time in therapeutic range, costs, and patient quality of life. While the current study by Whitehouse et al. alone is not yet sufficient to shift clinical practice away from warfarin at this time, it is an important and courageous first step toward generating supporting evidence that apixaban may be the most promising DOAC candidate for an RCT in the HM3 to date.