Advances in anticoagulation: patients with bioprosthetic heart valves

Anticoagulation, Heart valve, Bioprosthesis

Valve replacement is the only definitive treatment for patients with severe valvular heart disease. Mechanical valves are usually preferred over biological valves in younger patients because they are more durable but require lifelong anticoagulation with a vitamin K antagonist (VKA) to prevent thromboembolism. Bioprosthetic valves are associated with low rates of clinical thromboembolism, thereby obviating the need for long-term anticoagulation. However, newer imaging technologies (4D echocardiography, 4D computed tomography) have demonstrated subclinical valve thrombosis in 10–20% of patients within the first year of implantation, which may be a cause of structural valve deterioration. Accordingly, there is increasing interest in the optimal choice of antithrombotic therapy in these patients, particularly during the first 3 months after valve implantation when the risk of subclinical thromboembolism appears to be greatest.

In the absence of high-quality evidence from randomized trials, guidelines have given weak recommendations for the use of anticoagulants in patients with bioprosthetic valves. For patients with a surgical bioprosthetic valve, they suggest anticoagulation with VKA for 3–6 months, whereas for transcatheter valve replacement (TAVR), they suggest antiplatelet therapy. Guidelines recommend against the use of direct oral anticoagulants (DOACs) in patients with mechanical heart valves because in the REALIGN trial, dabigatran was associated with higher rates of both thromboembolism and bleeding than warfarin.¹ However, factor Xa inhibitors have not been tested in patients with mechanical valves and it is unclear whether the results with dabigatran in mechanical valves can be extrapolated to bioprosthetic valves.

Recent randomized trials (Table 1) have shed further light on the efficacy and safety of DOACs in patients with bioprosthetic heart valves. The first of these was GALILEO, published in 2020, which demonstrated in 1644 patients undergoing TAVR that the factor Xa inhibitor rivaroxaban 10 mg once daily (plus aspirin for the first 3 months) was less effective than aspirin (plus clopidogrel for the first 3 months) for prevention of death or thromboembolism.² Patients in the rivaroxaban group, however, had markedly lower rates of valve leaflet-motion abnormalities at 3 months (2.1% vs. 10.9%, P = 0.01), suggesting the potential for DOACs to prevent subclinical valve thrombosis.

The RIVER trial, also published in 2020, demonstrated in 1005 patients with a bioprosthetic mitral valve and atrial fibrillation (AF) (∼13.8% were taking antiplatelet therapy at baseline) that rivaroxaban 20 mg once daily was non-inferior to warfarin for prevention of death, major cardiovascular events or bleeding.³ However, only 189 patients in RIVER were enrolled within 3 months of surgery, and while clinical valve thrombosis was uncommon (1.04 vs. 0.62 per 100 patient-years), routine valve imaging was not performed.

ENAVLE was a much smaller trial (n = 220) evaluating anticoagulation in patients during the first 3 months after surgical bioprosthetic aortic valve replacement (AVR) or valve repair. Presented in 2020 and published in 2021, the results were consistent with those of RIVER in demonstrating that the factor Xa inhibitor edoxaban was non-inferior to warfarin for the prevention of death, clinical thromboembolic events, or asymptomatic intracardiac thrombosis.⁴

ENVISAGE-TAVI AF is the latest trial in this field. In 1426 patients with AF and TAVR (60% received antiplatelet therapy), edoxaban was non-inferior to warfarin for the prevention of the primary outcome of death from any cause, myocardial infarction, ischaemic stroke, systemic thromboembolism, valve thrombosis, or major bleeding, although it did increase major bleeding.⁵ All patients were randomized within 7 days of TAVR, thereby confirming the results of RIVER that DOACs are similarly effective to warfarin for the prevention of clinical thromboembolism early after bioprosthetic valve replacement. However, like RIVER and ENAVLE, the ENVISAGE-TAVI AF trial did not include routine valve imaging.

In parallel with efforts to evaluate the role of DOACs in patients with bioprosthetic heart valves, emerging evidence suggests that a new class of anticoagulant that targets coagulation factor XI (FXI) has the potential to replace VKAs for both mechanical and bioprosthetic heart valves. The promise of this new drug class was first suggested by the observation that patients with mild congenital FXI deficiency (FXI levels 10–20%) have low rates of both thromboembolism and bleeding.⁶ Because FXI inhibitors directly target the contact pathway, they selectively inhibit coagulation activation induced by exposure of blood to artificial surfaces, including the sewing rings and mechanical leaflets of prosthetic heart valves. Table 2 summarizes classes of FXI inhibitors that are in the most advanced stages of clinical evaluation.

Table 1

Randomized trials of anticoagulation for transcatheter or surgical bioprosthetic heart valves

Trial	Population	N	Intervention	Control	Outcomes
GALILEO 2020	Transcatheter AVR	1644	Rivaroxaban 10 mg once daily (+aspirin 100 mg once daily for the first 3 months)	Aspirin (+clopidogrel 75 mg once daily for the first 3 months)	Rivaroxaban + aspirin vs. aspirin + clopidogrel HR 1.35 (95% CI, 1.01–1.81) for primary outcome of death or first thromboembolic event, P = 0.04
RIVER 2020	Bioprosthetic MVR with AF	1005	Rivaroxaban 20 mg once daily	VKA	Edoxaban vs. VKA RMST difference 7.4 days (95% CI, −1.4 to 16.3); until death, major cardiovascular event, or major bleeding, P < 0.001 for non-inferiority
ENAVLE 2021	Surgical bioprosthetic AVR or valve repair	218	Edoxaban 60 mg once daily	VKA	Edoxaban vs. VKA risk difference -0.037 (95% CI, 0.0720 to –0.0014); for death, clinical thromboembolic events, or asymptomatic intracardiac thrombosis, P < 0.001 for non-inferiority
ENVISAGE-TAVI AF 2021	Transcatheter AVR with AF	1426	Edoxaban 60 mg once daily	VKA	Edoxaban vs. VKA HR 1.05 (95% CI, 0.85–1.31) for primary outcome of death from any cause, myocardial infarction, ischaemic stroke, systemic thromboembolism, valve thrombosis, or major bleeding, P = 0.01 for non-inferiority

AVR, aortic valve replacement; CI, confidence interval; HR, hazard ratio; MVR, mitral valve replacement; RMST, restricted mean survival time; VKA, vitamin K antagonist.

Table 2

Classes of factor XI inhibitors in the most advanced stages of clinical development

Class	Indications being evaluated	Agents	Mechanism of action	Route of administration	Onset	CYP 450 metabolism	Renal excretion
Antisense oligonucleotides	Total knee arthroplasty Haemodialysis	IONIS-FXIRx ISIS-FXIRx	Synthesized nucleic acid sequences which bind complementary mRNA and inhibit translation into FXI protein	Subcutaneous	Weeks	No	No
Monoclonal antibodies (IgG)	Total knee arthroplasty Haemodialysis Atrial fibrillation Cancer and indwelling central lines receiving chemotherapy	Osocimab Abelacimab Xisomab	Block activation and/or activity of activated FXI	Subcutaneous or intravenous	Hours to days	No	No
Small molecule inhibitors	Atrial fibrillation Post-myocardial infarction in patients receiving DAPT Post-stroke in patients receiving antiplatelet therapy	Asundexian Milvexian	Inhibit FXIa active site	Oral	Minutes to hours	Yes	Minimal (<20%)

CYP450, cytochrome P450; FXI, factor XI; FXIa, factor XIa activated; IgG, immunoglobulin G; mRNA, messenger ribonucleic acid.

The phase 2 ANT-005 TKA trial, published earlier this year, demonstrated in 412 patients undergoing total knee arthroplasty that a single post-operative intravenous dose of abelacimab, a FXI monoclonal antibody inhibitor, was effective for the prevention of venous thromboembolism and was associated with a low risk of bleeding.⁷ These results confirm the potential for drugs that target FXI to reduce thromboembolism without increasing bleeding,^8–10 thereby uncoupling thrombosis from haemostasis. So far, FXI inhibitors have not been evaluated in patients with prosthetic heart valves.

What have we learned from the recent randomized trials of anticoagulation in patients with bioprosthetic valves? First, the results confirm previous observational data that rates of clinical thromboembolism are low following a surgical or transcatheter bioprosthetic heart valve replacement. Second, DOACs given at therapeutic doses appear to be as effective as VKAs for prevention of clinical thromboembolism in these patients. Third, although the effect of DOACs on subclinical valve thrombosis remains incompletely studied, preliminary data from GALILEO suggest that they may be effective even when given at lower doses.

What are the implications of the latest data for clinical practice? Based on the current evidence, the choice of valve prosthesis (mechanical vs. bioprosthetic) should continue to be based on considerations of valve durability and the need for anticoagulation. For patients undergoing bioprosthetic valve replacement who require anticoagulation for another indication (e.g. AF), DOACs appear to be a reasonable choice, including in the first 3 months after valve implantation. For patients with a surgical bioprosthetic valve who do not have another indication for anticoagulation, the data favour anticoagulation over no anticoagulation for the first 3 months, but the evidence is not of high quality.

What are the implications for future research? Although the effects of low-dose rivaroxaban on subclinical thrombosis in the GALILEO trial are promising, this needs to be confirmed and we do not yet know whether prevention of subclinical valve thrombosis can improve long-term valve durability. If an improvement in valve durability can be achieved with DOAC anticoagulation, it may allow for greater use of bioprosthetic valves in younger patients. In the meantime, we await further trials of the FXI inhibitors to determine whether the potential to uncouple thrombosis from bleeding is realized, and even more importantly, whether they can provide effective, safer, and more convenient anticoagulation in patients with prosthetic heart valves.

Data availability statement: No new data were generated or analysed in support of this research.

Conflict of interest: SC has received speaker fees from BMS Pfizer and thrombosis fellowship funding from Sanofi Canada. JWE has received honoraria and/or research support from AZ, Bayer, BI, BMS, Janssen, Pfizer, Portola, and Servier.