Article Text

Download PDFPDF

Antithrombotic therapy in patients undergoing transcatheter aortic valve implantation
  1. Vincent Johan Nijenhuis1,
  2. Jorn Brouwer1,
  3. Lars Søndergaard2,
  4. Jean-Philippe Collet3,
  5. Erik Lerkevang Grove4,5,
  6. Jurrien Maria Ten Berg1
  1. 1 Department of Cardiology, St Antonius Hospital, Nieuwegein, The Netherlands
  2. 2 Department of Cardiology, Rigshospitalet, Copenhagen, Denmark
  3. 3 Action Study Group, Institut de Cardiologie de la Pitié-salpêtrière (APHP), Paris, France
  4. 4 Department of Cardiology, Aarhus University Hospital, Aarhus, Denmark
  5. 5 Department of Clinical Medicine, Faculty of Health, Aarhus University, Aarhus, Denmark
  1. Correspondence to Dr Jurrien Maria Ten Berg, Department of Interventional Cardiology, St Antonius hospital, Nieuwegein 3435 CM, The Netherlands; j.ten.berg{at}


This review provides a comprehensive overview of the available data on antithrombotic therapy after transcatheter aortic valve implantation (TAVI). In the absence of large randomised clinical trials, clinical practice is leaning towards evidence reported in other populations. Due to the greater risk of major bleeding associated with oral anticoagulation using a vitamin-K antagonist (VKA), antiplatelet therapy (APT) may be considered as the first-line treatment of patients undergoing TAVI. Overall, single rather than dual APT is preferred. However, dual APT should be considered in patients with a recent acute coronary syndrome (ie, within 6 months), complex coronary stenting, large aortic arch atheromas or previous non-cardioembolic stroke. Monotherapy with VKA should be considered if concomitant atrial fibrillation or any other indication for long-term oral anticoagulation is present. APT on top of VKA seems only reasonable in patients with recent acute coronary syndrome, extensive or recent coronary stenting or large aortic arch atheromas. A direct-acting oral anticoagulant may be considered if oral anticoagulation is indicated in the absence of contraindications. Initiation of VKA is indicated in clinical valve thrombosis, for example, with high transvalvular gradient, whereas the role of VKA in the case of subclinical leaflet thrombosis is currently uncertain.

  • antithrombotic
  • aspirin
  • clopidogrel
  • anticoagulant drugs
  • aortic bioprosthesis
  • transcatheter aortic valve replacement (TAVR)
  • transcatheter aortic valve implantation (TAVI)

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Transcatheter aortic valve implantation (TAVI) is a well-established treatment for severe symptomatic aortic valve stenosis in patients with increased surgical risk,1–7 yet the optimal antithrombotic regimen after TAVI is unknown. Currently available studies, limited by their underpowered and mainly observational nature, have not elucidated whether single or dual antiplatelet therapy (SAPT or DAPT) or oral anticoagulant (OAC) therapy [that is, vitamin-K antagonist (VKA) or direct-acting oral anticoagulant (DOAC)] after TAVI is most appropriate. Consequently, preferred therapy in clinical practice is highly variable.8 Some patients may require APT or OAC based on other indications than TAVI. Current TAVI guidelines are mainly based on expert opinion and recommend DAPT for the first 1 to 6 months followed by life-long low-dose aspirin, although in case of low bleeding risk, use of VKA may be reasonable.9 10 In patients with atrial fibrillation (AF) or another indication for OAC, a combination of OAC and aspirin or clopidogrel is usually recommended (figure 1).

Figure 1

Graphic summary of current guideline recommendations in the overall TAVI population based on the guidelines and Expert Consensus Statements of the ESC,9 and ACC/AHA.10 ACC, American College of Cardiology; AHA, American  Heart  Association; ASA, acetylsalicylic acid; DAPT, dual antiplatelet therapy; DOAC, direct-acting OAC; ESC, European Society of Cardiology; OAC, oral anticoagulation; SAPT, single antiplatelet therapy; TAVI, transcatheter aortic valve implantation; VKA, vitamin-K antagonist.

This review summarises currently available data and suggests a more individualised antithrombotic treatment strategy after TAVI based on risk factors for cardiovascular events and bleeding.

Antiplatelet therapy

Key messages

  • SAPT rather than DAPT may be considered when there is no need for OAC.

  • DAPT rather than SAPT may be considered in patients with a recent acute coronary syndrome (ie, within 6 months), complex coronary stenting, large aortic arch atheroma’s or previous non-cardioembolic stroke.

When there is no indication for OAC, SAPT may be considered post-TAVI to prevent cardiovascular events. Preliminary data suggest that DAPT offers no additional benefit compared with SAPT in the overall TAVI population. A meta-analysis of two small randomised clinical trials and two observational studies (n=672) showed no significant differences between DAPT and SAPT after TAVI with regard to net adverse events or its components at 30 days.11 There was, however, a trend towards a higher rate of bleeding in patients receiving DAPT versus aspirin monotherapy.

More recently, Mangieri et al showed that treatment with DAPT for approximately 5 months (n=331) compared with SAPT (n=108) did not result in different rates of net adverse clinical events and valve thrombosis at 1 year.12 A higher rate of major or life-threatening bleeding in DAPT (21%) than SAPT (8%) was found in a smaller single-centre study (n=144).13 In the Aspirin Versus Aspirin+ClopidogRel Following TAVI trial, randomisation to DAPT versus SAPT in 222 patients resulted in an increased risk of major or life-threatening bleeding events but no differences in the rate of myocardial infarction or stroke.14 The ongoing APT fOr Patients Undergoing TAVI trial (NCT02247128) is testing the safety and efficacy of SAPT versus DAPT and OAC+clopidogrel versus OAC monotherapy.15 The ongoing safety profile evaluation of TICagrelor Alone compared with a Combination of Lysine Acetylsalicylate-Clopidogrel in the Context of TAVI trial (NCT02817789) will test the 30-day safety profile of SAPT using ticagrelor. These and other ongoing randomised trials are depicted in table 1.

Table 1

Active clinical trials evaluating antithrombotic strategies in patients undergoing TAVI

Some comorbidities guide us to the use of APT rather than OAC based on previous research. Coronary artery disease is present in approximately 70% of patients with TAVI1–7 and this is associated with an increased risk of subsequent cardiovascular events.16 Approximately 20% of patients suffered a previous myocardial infarction and had undergone coronary stenting (figure 2).

Figure 2

Graphic summary of incidences of bleeding and ischaemic complications according to time. Values are plotted as combined mean of event rates and SE based on the incidences reported in large randomised controlled trials of intermediate and high-risk patients.1–7

In patients with coronary artery disease, APT rather than VKA is indicated given the reduced risk of major bleeding.17 DAPT rather than SAPT is often prescribed to prevent coronary events. Long-term DAPT reduces stent thrombosis and spontaneous myocardial infarction, although offset by an increased risk of clinically relevant bleeding. Overall, a mortality benefit with long-term DAPT has not been shown.18 Thus, the DAPT duration in these settings should follow current guidelines.19

APT is also indicated in patients with symptomatic peripheral artery disease to reduce the risk of cardiovascular events.20 Peripheral artery disease is present in approximately 40% of patients undergoing TAVI1–7 (figure 2) and routine multiple detector CT (MDCT) imaging plays a central role in the patient selection and procedure route planning.

In patients with aortic arch atheroma, APT is superior to VKA. The Aortic Arch Related Cerebral Hazard trial found that patients with an ischaemic stroke and aortic arch plaques treated with DAPT had a lower rate of vascular death compared with VKA alone.21 We, therefore, recommend the use of APT in these patients.22

Extracranial carotid artery stenosis is present in approximately 30% of patients undergoing TAVI23 and is associated with an increased risk of stroke after TAVI.24 In these patients, SAPT is recommended as primary prevention for ischaemic cardiovascular events.20 APT rather than VKA is recommended as secondary prevention based on the comparable efficacy but higher rates of major bleeding associated with VKA compared with SAPT.22 DAPT is preferred in patients with a recent prior stroke without AF for a maximal duration of 3 months, whereas SAPT is preferred thereafter.

Oral anticoagulation

Key messages

  • OAC monotherapy is recommended if concomitant AF or any other indication for long-term OAC is present.

  • Dual therapy with OAC plus SAPT may be considered if there is an indication for OAC and a recent acute coronary syndrome (ie, within 6 months), complex coronary stenting or large aortic arch atheromas.

  • DOAC may be considered if there is an indication for OAC and no contraindications.

AF is present in more than one-third of patients undergoing TAVI1–7 (figure 2). Additionally, clinical new-onset AF after TAVI occurs in up to 36%.1–7 In a subanalysis of the Nordic Aortic Valve Intervention trial, new onset AF was registered in approximately 80% of patients after TAVI when employing reveal implants.25 New onset AF is associated with an increased risk of cardiovascular events, mortality and prolonged hospital stay.24 26 27

AF is a known predictor of stroke and systemic embolism, and the need for OAC places patients with TAVI at a higher risk of bleeding events postprocedure and during follow-up.24 When a bleeding complication occurs in patients with concomitant AF, the observed 1-year mortality after TAVI is increased to 50%.28 Therefore, the assessment of optimal antithrombotic therapy in these patients should be carefully considered. Given the risk of bleeding, additional APT in these patients should not be routinely administered but instead be reserved for patients with an absolute indication such as recent coronary stenting.

So far, VKA in patients undergoing TAVI has only been tested in those with AF. In patients with AF, VKA is indicated as it is more effective than SAPT and DAPT with regard to primary stroke prevention with an acceptable risk of bleeding.9 10 In patients without AF, the ongoing DAPT versus OAC for a Short Time to Prevent Cerebral Embolism After TAVI trial (NCT01642134) compares VKA with DAPT for the prevention of cerebral thromboembolism by the detection of new areas of cerebral infarction by MRI 3 months after TAVI.

APT on top of VKA is unlikely to confer any added clinical benefit after TAVI. The addition of APT to VKA in patients with TAVI is associated with similar stroke or death rates, but an increased rate of major and/or life-threatening bleeding.29 In another study, VKA monotherapy resulted in a lower risk of the combined endpoint of post-procedural mortality, stroke, embolism and major bleeding.30

In patients with AF and coronary artery disease, additional APT does not seem beneficial. Additional APT did not reduce the risk of recurrent coronary events or stroke but increased the risk of clinically relevant bleeding events.31 Therefore, in patients with TAVI, with stable coronary artery disease and AF, VKA monotherapy seems an appropriate choice, unless recent coronary stenting necessitates APT. Similarly, the addition of APT does not seem beneficial in patients with AF and peripheral artery disease.20

The benefit of DOAC in patients undergoing TAVI remains to be investigated. The Global Study compares a rivAroxaban-based Antithrombotic Strategy with an antipLatelet-based Strategy After Transcatheter aortIc vaLve rEplacement to Optimise Clinical Outcomes trial (NCT02556203) compares rivaroxaban in combination with aspirin for 3 months followed by rivaroxaban monotherapy versus DAPT for 3 months followed by aspirin monotherapy in patients without an indication for OAC.32 This study has been prematurely terminated because rivaroxaban 10 mg daily on top of aspirin for the first three postprocedural months followed by monotherapy with rivaroxaban was associated with an increase in the primary efficacy outcome of death or first adjudicated thromboembolic event (11.4% vs 8.8%), primary safety outcome of bleeding (4.2% vs 2.4%) and all-cause death (6.8% vs 3.3%).

The ongoing Anti-Thrombotic Strategy After Trans-Aortic Valve Implantation for Aortic Stenosis trial (NCT02664649) tests the safety and efficacy of apixaban compared with VKA in patients with an indication for OAC, and to DAPT/SAPT in patients without an indication for OAC. The ENVISAGE AF trial (NCT02943785) compares edoxaban with VKA in patients with AF and an indication for OAC (table 1).

Leaflet thrombosis

Key messages

  • VKA therapy should be considered when there is evidence for clinical leaflet thrombosis.

  • The role of OAC in the case of subclinical leaflet thrombosis is currently uncertain.

Clinical valve thrombosis

The risk of clinical (obstructive) transcatheter valve thrombosis is low, being reported in up to 3% of cases.33 34 Thus far, at least 75 cases of clinical transcatheter valve thrombosis have been described. Of these, only 21 (28%) occurred within the first 3 months suggesting that a longer period of antithrombotic therapy may be warranted. Almost none of the valve thrombosis occurred while on OAC, and valve function was restored after OAC therapy. Therefore, VKA initiation should be considered when there is evidence of clinical leaflet thrombosis.

Subclinical leaflet thrombosis

Subclinical leaflet thrombosis is a common finding on MDCT.35 Hypoattenuated leaflet thickening (HALT) and hypoattenuation affecting motion are two stages of the same phenomenon, with leaflet thickening affecting leaflet motion at a more advanced stage. These signs of leaflet thrombosis of one or more cusps have been found in approximately 15% of cases.36–40

Leetmaa et al showed the presence of leaflet thrombosis as assessed by MDCT in 4% of patients, of whom 40% did not receive DAPT.40 In a study by Makkar et al, 21% patients showed leaflet thrombosis, and therapeutic VKA was associated with lower rates of reduced leaflet motion than DAPT, whereas no or subtherapeutic VKA was associated with 92% and DOAC with 8% of cases.36 All patients in whom therapeutic VKA was initiated had a recovery to normal leaflet function.

In the RESOLVE and SAVORY registries,41 subclinical leaflet thrombosis occurred more often in patients receiving APT than OAC (15% vs 4%) with no difference between VKA and DOAC and resolved in all cases with VKA administration. In the FRANCE-TAVI registry (n=12 804), OAC was associated with a lower rate of bioprosthetic valve dysfunction,42 defined as an increased prosthetic gradient ≥10 mm Hg as a possible surrogate of subclinical leaflet thrombosis. However, subclinical leaflet thrombosis may be a temporal dynamic finding as it can appear both early and late after valve implantation, as well as resolve without a change in antithrombotic therapy.35

In conclusion, OAC is associated with a lower incidence of clinical and subclinical leaflet thrombosis than APT. Moreover, OAC is effective in the reduction of leaflet thickening thus providing an additional argument that leaflet thickening is a thrombotic process. It is uncertain whether subclinical leaflet thrombosis may impact the postprocedural outcome and early structural valve deterioration. Registry data suggest that this phenomenon may be associated with a higher prevalence of transient ischaemic attack, but not with stroke.41 However, these findings must be interpreted with caution due to a temporal separation between the MDCT and the clinical event, and prospective clinical studies employing more sensitive measures of cerebral infarctions are warranted (ie, brain MRI). Similarly, there are currently no data supporting a potential negative impact of subclinical leaflet thrombosis on the durability of bioprosthetic aortic valves. Despite higher rates of leaflet thickening in transcatheter as compared with surgical heart valves,41 5-year echocardiographic follow-up results of the large randomised trials did not show more frequent valve degeneration or higher transvalvular gradients in transcatheter as compared with surgical heart valves. Therefore, any recommendations on routine screening for HALT and use of a preventive VKA strategy would be premature.

Bleeding risk

Key messages

  • Left atrial appendix closure in AF may be considered in selected patients with contraindications for OAC or high bleeding risk.

It is recognised that several risk factors predisposing to bleeding are also risk factors for stroke (table 2). Accordingly, assessment of additional more specific risk factors may improve the assessment of bleeding risk and facilitate the choice of optimal antithrombotic therapy.

Table 2

Risk factors for bleeding and ischaemic events

During the TAVI procedure, patients are at increased risk of bleeding in particular when non-transfemoral access is required.43 In patients undergoing transfemoral TAVI, the most important predictors of bleeding include the introducer sheathtofemoral artery ratio and femoral artery calcium score.44 In these patients at increased bleeding risk, careful temporary reduction of the intensity of antithrombotic therapy may be considered by means of either stopping VKA, delaying initiation of APT or starting APT without a loading dose.

Long-term bleeding risk may specifically be affected by frailty. Frailty is characterised by weakness, a decreased physiological reserve and less adaptation to stressors such as acute illness or trauma. It is found in approximately half of the patients undergoing TAVI.45 In frail individuals, fibrin generation is reduced46 and the risk of major bleeding is increased independently of age.47 Anticoagulation is often avoided given the perceived fall risk that has been associated with bone fractures, bleeding events and mortality.48 Patient’s frailty and risk of falling should be reassessed during follow-up and antithrombotic treatment should be adjusted accordingly.

Furthermore, the long-term bleeding risk may be assessed with the use of biomarkers, such as a prolonged ADP closure time.49 The incremental values of such markers should be further investigated to possibly guide interventions to paravalvular leak closure or adaptations in the antithrombotic regimen.

In patients with AF at high risk of bleeding or those who previously suffered a major bleeding or an ischaemic stroke despite well-regulated OAC, a concomitant left atrial appendage closure may be considered. A pilot study showed that concomitant TAVI and left atrial appendage closure are feasible and safe.50 The WATCHMAN-TAVR trial (NCT03173534) compares the simultaneous left atrial appendage closure by Watchman implantation and TAVI to TAVI with medical management for AF in 312 patients with a 2-year follow-up. Until results confirm the safety and efficacy of such an approach, no recommendation can be made. However, staged procedures may be reasonable to perform. A proposed individualised treatment scheme based on the evidence described above is shown in figure 3.

Figure 3

Potential antithrombotic treatment options for prevention of thromboembolism after TAVI. Bleeding risk is estimated using current available risk scores. such as HAS-BLED, and clinical characteristics such as described in table 2. AF, atrial fibrillation; ACS, acute coronary syndrome; DAPT, dual antiplatelet therapy; NCE, non-cardioembolic; PCI, percutaneous coronary intervention; SAPT, single antiplatelet therapy; TAVI, transcatheter aortic valve implantation; VKA, vitamin-K antagonist.


Acknowledging the ongoing population shift of TAVI to patients at lower surgical risk and younger age, the decision for bioprosthetic rather than mechanical aortic valves becomes more relevant and the optimal antithrombotic regimen may change. An important reason to choose a bioprosthetic aortic valve is to avoid the need for long-term VKA. However, if long-term VKA therapy is needed, mechanical valves may be a better option than transcatheter or surgical bioprosthetic valves because of better durability. Therefore, the clinical consequences of subclinical leaflet thrombosis after TAVI and the role of (possibly short term) VKA or DOAC for its prevention should be tested.

Starting points for research and research perspectives

  • Testing the optimal APT regimen (aspirin, clopidogrel, ticagrelor).

  • Testing the optimal DOAC regimen, for example, comparing reduced dose DOAC+SAPT to DOAC monotherapy and SAPT.

  • Effect of subclinical leaflet thrombosis on ischaemic markers, for example, silent cerebral infarctions on MRI.

  • Clinical consequences of subclinical leaflet thrombosis.


Optimal antithrombotic therapy needs to address both the early device-related risks and more constant patient-related thrombotic risk. Thus, choosing the optimal antithrombotic strategy should not only be based on the transcatheter valve used and the procedure performed, but also on the individual patient’s previous history and risk factors for bleeding and thrombosis. In the absence of large randomised clinical trials, we are leaning towards evidence reported in other than the TAVI population. Awaiting the results of ongoing trials, APT rather than VKA may be considered as a primary treatment of patients undergoing TAVI given the greater risk of major bleeding with VKA. In patients with AF, VKA is the treatment of choice. Additional APT seems only reasonable in patients with recent ACS, extensive or recent coronary stenting or large aortic arch atheromas defined as ≥4 mm in thickness. Initiation of OAC is indicated in clinical valve leaflet thrombosis, whereas the role of OAC in the case of subclinical leaflet thrombosis is currently uncertain.



  • Contributors Writing of the manuscript: VN, JB, LS, J-PC, ELG, JB. Critical revision of the manuscript for intellectual content: VN, JTB, LS, J-PC, EG, JB. Authors responsible for the overall content as guarantors: VN, JB.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests The authors have no direct conflicts related to the topic, but report the following general conflicts: ELG has received speaker honoraria or consultancy fees from AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Pfizer, MSD and Roche. JMB has received advisory/consulting/speakers fees from AstraZeneca, Eli Lilly, Daiichi Sankyo, the Medicines Company, Accumetrics, Boehringer-Ingelheim, BMS, Pfizer, Bayer, ferrer, and research grants from ZonMw and AstraZeneca.

  • Provenance and peer review Not commissioned; externally peer reviewed.