Discussion
In this study, we have found that: (1) there is large variance in the antiplatelet regimens used by different TAVI operators, with the majority of operators relying on personal or institutional policies rather than guideline recommendations; (2) compared with DAPT, aspirin monotherapy demonstrated similar protection against stroke; and (3) if DAPT is preferred, 1 month appears to be sufficient.
The meta-analysis presented is, by some margin, the largest addressing the question of the optimal antiplatelet strategy following TAVI. In this field, where adequately sized, high-quality RCTs are currently lacking, this represents the most comprehensive analysis to date to inform clinical practice.
In addition to providing the largest and most contemporary data available, our study has further unique aspects. First, we have performed a structured survey that highlights the patterns of antithrombotic use by TAVI operators in clinical practice, and the rationale behind this. Furthermore, we have gone on to perform a metaregression using antiplatelet strategy as a moderator. This additional analysis confirms that prolonged DAPT had no effect on clinical outcomes, and helps give clinicians further confidence in limiting the duration of DAPT or using a single antiplatelet strategy for their patients.
Disparity between guidelines and clinical practice
Current guideline recommendations on antithrombotic therapy after TAVI are not uniform, and the strength of recommendations is not strong. The 2014 and 20178 9 American Heart Association/American College of Cardiology (ACC) guidelines on valvular heart disease suggest aspirin lifelong plus clopidogrel for 6 months ‘may be reasonable’, corresponding to a strength of recommendation class of IIb. The level of evidence is classified as level C, indicating the lowest possible level of certainty (very limited populations evaluated, only consensus opinion of experts, case studies or standards of care). The 2017 ACC Expert Consensus Document45 on the care of patients after TAVI states that the current standard antithrombotic therapy after TAVI should be clopidogrel 75 mg orally daily for 3–6 months with oral aspirin 75–100 mg daily lifelong. The European Society of Cardiolog y guidelines on valvular heart disease7 do not make a specific recommendation for the duration of DAPT after TAVI, but simply state that a combination of low-dose aspirin and a thienopyridine should be used early after TAVI, followed by aspirin or a thienopyridine alone. The definition of ‘early’ is not provided, leaving this open to interpretation by clinicians. Position statements from the Canadian Cardiovascular Society (CCS)46 (DAPT for 1–3 months) and American College of Cardiology Foundation (ACCF)/American Association for Thoracic Surgery (AATS)/Society for Cardiovascular Angiography and Interventions (SCAI)/Society of Thoracic Surgeons (STS)47 (DAPT for 3–6 months) give similar recommendations.
Clinicians treating patients after TAVI are therefore provided with guideline recommendations that are not uniform across guideline bodies, are not specific with regard to the duration of DAPT recommended within individual guidelines and are made with the lowest levels of recommendation and weakest strength. This is a reflection of the paucity of evidence available in the field, and has likely contributed to the heterogeneous patterns of antithrombotic therapy clinicians are using.
For the first time, in this study, we present contemporary, real-world clinical practice data with regard to antiplatelet prescribing following TAVI, and the rationale behind treatment choices made by clinicians. This has provided an insight into how treating clinicians chose to interpret the available data, along with guideline recommendations and their own clinical judgement to inform the choice of antithrombotic therapy for patients undergoing TAVI.
The majority of the clinicians surveyed do not base their decisions on antithrombotic regimens after TAVI on guideline recommendation (only 13.3%); rather, the vast majority base their decisions on personal preference or local institutional policy. This may be a reflection of the variance in recommendations between different guideline bodies, as well as the weak strength of recommendation and low class of evidence. The guideline recommendations are in turn a reflection of the absence of strong randomised data in the field. The most commonly used duration of DAPT in our structured survey was 3 months (57% of respondents), while only 5% of centres used a shorter term policy (1 month).
Only 11.1% of operators choose their antithrombotic regimen according to the valve type inserted. The recommendation for the Edwards SAPIEN series of valves is 6 months of DAPT followed by aspirin lifelong; this is because this was the regimen used in the Placement of Aortic Transcatheter Valves (PARTNER) series of trials.3 After the implantation of a CoreValve, the recommendation is for 3 months of DAPT then aspirin lifelong as this was the suggested strategy in the CoreValve trials.10
Optimal duration of DAPT
The majority of centres used DAPT after TAVI. The most common duration was 3 months with only 5% of centres using 1-month DAPT. This study demonstrates that there was no additional benefit of durations of DAPT longer than 1 month with regard to the prevention of stroke or death.
There was no difference in bleeding rate between DAPT durations in this analysis. This may be a reflection of the heterogeneous definitions of bleeding between the studies.
Other studies have demonstrated the clear bleeding risk of prolonged DAPT.34 Although these studies did not include patients undergoing TAVI, their findings are likely to be applicable to the TAVI population given the comparatively high prevalence of renal impairment, underlying anaemia, frailty and other comorbidities that increase the propensity to bleed. It is therefore reasonable to conclude that should DAPT be preferred, the duration should be limited to 1 month.
Aspirin monotherapy is equivalent to DAPT
Guidelines recommend DAPT after TAVI, but as described above these are empirical recommendations based largely on expert consensus view. In the absence of RCT data, the recommendation for DAPT over aspirin monotherapy has been founded on extrapolation of data advocating DAPT following implantation of drug-eluting coronary stents. However, there are fundamental differences in the thrombotic tendencies of TAVI valves and coronary stents, which limit the acceptability of this assumption. Important differences include the increased size and the bioprosthetic nature of TAVI valves. Furthermore, the advanced age, frailty and comorbidity of the current cohort of patients undergoing TAVI make bleeding complications more likely.
This analysis suggests DAPT may not be necessary. Aspirin monotherapy was found to be equivalent to all durations of DAPT in terms of stroke prevention and death. This is consistent with two small randomised trials26 42 (total of 199 patients) which have compared DAPT with aspirin monotherapy, as well as two non-randomised studies28 33 (total of 463 patients). When considered alongside our analysis of 11 781 patients, there appears to be a consistent signal that aspirin monotherapy is as effective as DAPT at preventing stroke.
The recently published randomised controlled ARTE trial (Aspirin Versus Aspirin+Clopidogrel Following TAVI) was prematurely halted following an excess of major and life-threatening bleeds in the DAPT allocated group. Importantly, no increased risk of stroke was observed in those patients randomised to aspirin monotherapy. The findings therefore from this small trial comprising just 222 patients would concur with the conclusions of this much larger meta-analysis, specifically that DAPT does not confer additional advantage over aspirin monotherapy.48
Potential role of new oral anticoagulants
Recent data have shown that subclinical valve leaflet thrombosis is more common after TAVI than surgical aortic valve replacement with bioprosthetic valves,49 and that this may lead to an increased risk of stroke. Therefore, there is interest in the potential role of oral anticoagulants, particularly non-vitamin K oral anticoagulants (NOAC) for antithrombotic therapy after TAVI. The role of NOACs after TAVI is being evaluated in prospective RCTs (ClinicalTrials.gov Identifier: NCT02664649) and these data may inform future guideline recommendations regarding antithrombotic therapy after TAVI.
Limitations
To limit the possibility of a false positive result, we have used a random effects analysis. However, it remains unclear how the number of studies, the relative weights of each study, the extent of heterogeneity between studies and the potential for aggregation bias all influence the probability of any metaregression producing false positive or false negative results. We assessed publication bias using a Funnel plot, but accept that this does not completely exclude the possibility of publication bias. Therefore, as with any other meta-analysis, any conclusions from this study should be interpreted with this uncertainty in mind. The number of patients treated with aspirin monotherapy was much lower than those treated with DAPT, probably as a reflection of the current guideline recommendations. One could argue that this may limit the comparability of these cohorts. However, the event rates seen in the monotherapy groups are similar to the DAPT groups, and furthermore short durations of DAPT were as safe and effective as longer durations of DAPT; this provides support to the idea that aspirin monotherapy may be sufficient.
There was no uniform definition of bleeding. This may have masked any effect longer DAPT durations may have had on bleeding. A further explanation for the lack of effect of DAPT on bleeding may lie in the fact that some TAVI trials have excluded patients with recent bleeding events or at particularly high risk for bleeding. This incurs an inevitable selection bias which may have affected our results. The lack of effect of DAPT on bleeding should therefore be interpreted with caution.