Aspirin Alone Versus Dual Antiplatelet Therapy After Transcatheter Aortic Valve Implantation: A Systematic Review and Patient‐Level Meta‐Analysis

Nonstandard Abbreviations and Acronyms

Since its introduction, transcatheter aortic valve implantation (TAVI) has evolved as the alternative treatment for surgery in patients with severe symptomatic aortic valve stenosis.¹, ², ³, ⁴, ⁵, ⁶ Newer device technologies and better implantation techniques have largely improved the outcomes of TAVI over the past years. One aspect that has not been sufficiently investigated yet is the optimal antithrombotic treatment after TAVI. What we do know is that both bleeding and thromboembolic risk are high in patients undergoing TAVI. However, the optimal antithrombotic strategy to minimize both these risks is unknown in patients without a long‐term indication for oral anticoagulation after TAVI. The latest guidelines recommend dual antiplatelet therapy (DAPT) with aspirin and clopidogrel for the first 3 to 6 months, followed by aspirin alone lifelong.⁷, ⁸ Small underpowered randomized trials suggested that aspirin alone after TAVI reduced the incidence of bleeding, and did not increase thromboembolic events compared with DAPT.⁹, ¹⁰, ¹¹ The recently published POPular TAVI cohort A (Antiplatelet Therapy for Patients Undergoing Transcatheter Aortic Valve Implantation) trial in patients without a long‐term indication for oral anticoagulation after TAVI confirmed a lower incidence of bleeding and no increase of thromboembolic events with aspirin alone compared with DAPT.¹² The POPular TAVI trial was powered for bleeding events and a composite end point including bleeding and thromboembolic events but not for thromboembolic events alone. In order to increase statistical power, we performed a pooled individual patient‐level meta‐analysis from randomized clinical trials, to further evaluate the efficacy of aspirin alone compared with DAPT for preventing both thromboembolic and bleeding events after TAVI in patients without an indication for oral anticoagulation.

Methods

The data that support the findings of this study are available from Dr. J.M. ten Berg (jurtenberg@gmail.com) upon reasonable request.

Search Strategy

A systematic review of the published data on antithrombotic therapy after TAVI in patients without an established indication for oral anticoagulation was conducted, in accordance with the guidance and reporting items specified in the Preferred Reported Items for Systematic Reviews and Meta‐Analysis (PRISMA) statement.¹³ A literature search in the electronic databases PubMed, EMBASE, and Cochrane was last performed on September 1, 2020. Additionally, conference proceedings and abstracts, Clinicaltrials.gov, and the bibliography of review articles, meta‐analyses, or original studies identified by the literature search were screened for other eligible studies. The search included the following terms: “TAVI,” “TAVR,” “transcatheter aortic valve implantation,” “transcatheter aortic valve replacement,” “antiplatelet therapy,” “single antiplatelet therapy,” “dual antiplatelet therapy,” “SAPT,” “DAPT,” “aspirin,” “clopidogrel,” “antiplatelet,” and “antithrombotic therapy” (Complete search is provided in Data S1). Only prospective, randomized clinical trials were selected for inclusion in this meta‐analysis; other study designs were excluded. Ultimately, 4 prospective randomized clinical trials investigating aspirin alone compared with aspirin with clopidogrel after TAVI were included in the meta‐analysis: (1) the POPular TAVI trial; (2) the ARTE (Aspirin Versus Aspirin Plus Clopidogrel Following Transcatheter Aortic Valve Implantation) trial; (3) the SAT‐TAVI (Single Antiplatelet Therapy for TAVI) trial; and (4) the Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial from Ussia et al.⁹, ¹⁰, ¹¹, ¹² Access to the anonymized patient‐level data of all eligible trials was granted by the principal investigator of each trial. Data on baseline characteristics, procedural characteristic, and outcomes were verified using the original publications, and pooled into 1 single database. Two independent reviewers (J.B. and V.J.N.) assessed the risk of bias for all included trials using the Cochrane Collaboration's tool for assessing risk of bias in randomized trials (available in Tables S1 through S4).¹⁴ All included trials were approved by the local ethics committee and institutional review board of each participating center and all patients provided written informed consent.

Definitions and End Points

The primary outcome of this meta‐analysis was the composite of all‐cause mortality, stroke, or myocardial infarction at 30 days and 3 months. The co‐primary outcome was the composite of all‐cause mortality, major or life‐threatening bleeding, and stroke or myocardial infarction at 30 days and 3 months. The secondary outcomes were all‐cause mortality, cardiovascular mortality, major or life‐threatening bleeding, stroke and myocardial infarction at 30 days and 3 months. All outcomes were defined according to the VARC‐2 (Valve Academic Research Consortium‐2).¹⁵

Data Analysis

The databases with individual patient‐level data of the 4 trials were combined into 1 pooled database. Data on the outcomes at 30 days were available for all included trials, the outcomes at 3 months including time‐to‐event information were available for the POPular TAVI trial, the ARTE trial, and the Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial.⁹, ¹¹, ¹² For this meta‐analysis, the outcomes were analyzed using a 2‐step approach. First, the relative risk measures (ie, odds ratios or hazard ratios) and their standard errors for the association of aspirin alone versus aspirin plus clopidogrel were obtained for the primary and secondary outcomes. Second, these estimates were pooled using inverse variance–based fixed‐effect meta‐analysis. Estimates of random‐effect meta‐analysis were also depicted in the forest plots. Between‐study variance was estimated by Restricted Maximum Likelihood approach. Heterogeneity among studies was estimated by χ² test and the I² statistics. For the primary outcomes, time‐to‐event data at 3 months of follow‐up were available in 3 studies (POPular TAVI, ARTE, and Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation), hence cumulative incidence graphs for aspirin monotherapy versus aspirin plus clopidogrel were constructed. The pooled hazard ratios shown on the cumulative incidence plots were obtained from stratified Cox‐regression models with the individual studies serving as the strata variable, which is similar to the 2‐stage fixed‐effect meta‐analysis.

Prespecified subgroup analyses of the primary outcomes at 30 days and 3 months were performed, which included sex, age (≥80 versus <80), renal failure (estimated glomerular filtration rate ≤60 versus >60), body mass index (≥25 versus <25), approach (transfemoral versus other), and prior stroke. A P value of <0.05 was considered statistically significant. All statistical analyses were performed using R version 3.3.3. (R Foundation for Statistical Computing, Vienna, Austria).

Results

Study Selection

We identified 402 reports with our search. After removal of duplicates, 302 studies were screened for eligibility, after screening for title and abstract 292 studies were excluded (ie, not topic of interest N=219, review or meta‐analysis N=73), and after full text screening 10 additional studies were excluded because of a nonrandomized design. The 4 remaining trials were included in this meta‐analysis (Figure 1). The included trials were (1) the POPular TAVI trial, N=665; (2) the ARTE trial, N=222; (3) the SAT‐TAVI trial, N=120; and (4) the Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial, N=79.⁹, ¹⁰, ¹¹, ¹² All trials randomly assigned patients in a 1:1 ratio, and in the 2 multicenter trials (POPular TAVI and ARTE) randomization was also stratified by participating center. Both the SAT‐TAVI trial and Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial did not describe the randomization tool used for randomization. Patients were all analyzed according to the intention‐to‐treat principle and clinical outcomes were prespecified in the trial protocols and (except for the Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial) adjudicated by an independent blinded clinical end point committee. The risk of bias was considered low for the 3 largest trials and unclear for the smallest trial (Tables S1 through S4).

• Download figure
• Download PowerPoint

Study Characteristics

All patients were considered suitable for TAVI, as determined by a dedicated heart team of each center. The main trials' exclusion criteria included a long‐term indication for oral anticoagulation, previous coronary artery stenting requiring dual antiplatelet therapy at the time of TAVI, major bleeding within 3 months before TAVI, cerebrovascular accident or transient ischemic accident in the past 6 months, and an allergy or intolerance to aspirin or clopidogrel (or ticlopidine, only for the SAT‐TAVI trial). During the procedure, unfractionated heparin was administered with an aimed activated clotting time of >250 seconds.

The aspirin‐alone group received aspirin at a dose of 80 to 160 mg daily for the entire trial duration. In patients not on aspirin before TAVI, aspirin was initiated within 24 hours before the TAVI procedure. Patients randomized to DAPT received aspirin at a dose of 80 to 160 mg daily with clopidogrel at a dose of 75 mg daily (or ticlopidine 500 mg twice daily) for 3 to 6 months, followed by aspirin alone (80–160 mg daily) for the entire trial duration. Aspirin was initiated within 1 day before TAVI, in patients not already on aspirin therapy. An initial loading dose of 300 mg of clopidogrel was administered within 24 hours before TAVI in patients undergoing transfemoral TAVI and within 24 hours after TAVI in those with a nontransfemoral approach, followed by 75 mg daily for 3 to 6 months, except for the SAT‐TAVI trial where it was started directly after the procedure in all patients.¹⁰

Baseline and Procedural Characteristics

A total of 1086 randomized patients across the 4 included randomized trials were included in this meta‐analysis. Of the included patients, 542 were randomized to aspirin alone and 544 to DAPT (or ticlopidine) after TAVI. The pooled baseline characteristics and the distribution of all characteristics for each separate trial are shown in Table 1, including missing data notification of each characteristic. The pooled mean age was 79.9±6.8 years and 49.7% of the patients were women. The median Society of Thoracic Surgeons score was lower in the POPular TAVI trial, compared with the other trials (2.5% versus 4.5% and 5.1%). Procedural characteristics are shown in Table S5. Most TAVI procedures were performed using the transfemoral approach (86.9%) with a balloon‐expandable device (59.8%).

Primary and Secondary Outcomes at 30 Days

The composite of all‐cause mortality, major or life‐threatening bleeding, stroke, and myocardial infarction at 30 days occurred in 10.3% of patients in the aspirin‐alone group and 14.7% of patients in the DAPT group (odds ratio [OR], 0.67; 95% CI, 0.46–0.97, P=0.034). The second primary composite outcome, including all‐cause mortality, stroke, or myocardial infarction, occurred in 5.5% of patients receiving aspirin alone and in 6.6% of patients receiving DAPT (OR, 0.83; 95% CI, 0.50–1.38, P=0.47) at 30 days (Table 2 and Figure 2).⁹, ¹⁰, ¹¹, ¹² Major bleeding and the combination of major or life‐threatening bleeding occurred less in the aspirin‐alone group, compared with the DAPT group (2.4% versus 5.9%, 95% OR, 0.40; CI, 0.21–0.78, P=0.007, and 5.4% versus 10.1%, 95% OR, 0.51; CI, 0.32–0.82, P=0.005, respectively) (Table 2 and Figures S1 and S2). The individual incidences of life‐threatening bleeding (3.0% versus 4.2%, 95% OR, 0.76; CI, 0.39–1.51, P=0.44), stroke (2.8% versus 3.5%, 95% OR, 0.79; CI 0.39–1.60, P=0.51), myocardial infarction (0.7% versus 1.5%, 95% OR, 1.35; CI, 0.30–6.08, P=0.70), death (2.6% versus 3.1%, 95% OR, 0.82; CI, 0.40–1.72, P=0.61) and cardiovascular death (2.4% versus 3.1%, 95% OR, 0.76; CI, 0.36–1.61, P=0.47) were numerically lower in the aspirin‐alone group but not statistically different (Table 2 and Figures S3 through S7). Results of the prespecified subgroup analyses for both primary outcomes are shown in Figures S8 and S9, which did not show any evidence for effect‐modification according to sex, age, body mass index, renal function, TAVI approach, and history of stroke.

• Download figure
• Download PowerPoint

Primary and Secondary Outcomes at 3 Months

Three‐month follow‐up was only available for the POPular TAVI trial, the ARTE trial, and the Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial, which comprised a total of 966 patients: 481 randomized to aspirin alone and 485 patients to DAPT. At 3 months, a first primary composite outcome event was observed in 11.0% of patients receiving aspirin alone and in 16.5% of patients receiving DAPT (hazard ratio [HR], 0.66; 95% CI, 0.47–0.94, P=0.02). A second primary composite outcome event was observed in 6.9% of patients in the aspirin‐alone group and in 8.5% of patients in the DAPT group (HR, 0.82; 95% CI, 0.52–1.29, P=0.39) (Table 3, Figure 3, and Figure S10). The incidences of the secondary outcomes are listed in Table 3. Both major bleeding and the combination of major or life‐threatening bleeding were significantly less frequent in the aspirin‐alone group, compared with the DAPT group (2.3% versus 6.4%, 95% OR, 0.35; CI, 0.17–0.71, P=0.004, and 4.8% versus 10.5%, 95% OR, 0.43; 95% CI, 0.26–0.72, P=0.001, respectively) (Figures S11 and S12). As for the 30‐day outcomes, the individual incidences of life‐threatening bleeding (2.5% versus 4.1%, 95% OR, 0.67; CI, 0.31–1.44, P=0.30), stroke (3.1% versus 3.9%, 95% OR, 0.80; CI, 0.39–1.61, P=0.52), myocardial infarction (0.8% versus 1.9%, 95% OR, 1.01; CI, 0.25–4.07, P=0.99), death (3.7% versus 4.3%, 95% OR, 0.86; CI, 0.44–1.66, P=0.65), and cardiovascular death (3.3% versus 4.1.%, 95% OR, 0.80; CI, 0.40–1.59, P=0.52) at 3 months were numerically lower in the aspirin‐alone group as compared with the DAPT group, but these differences were not statistically significant (Figures S13 through S17). The results of the prespecified subgroups are shown in Figures S18 and S19. Heterogeneity between the included trials in treatment effects was low for all primary and secondary outcomes (I²<25%).

• Download figure
• Download PowerPoint

Discussion

In this patient‐level meta‐analysis of all available prospective randomized trials, we investigated the effect of aspirin alone compared with DAPT after TAVI on thromboembolic and bleeding complications in patients without an indication for oral anticoagulation. In summary, aspirin alone significantly reduced the rate of the composite of major and life‐threatening bleeding or thromboembolic events (including all‐cause mortality, major and life‐threatening bleeding, stroke, or myocardial infarction), compared with DAPT. The difference was caused by a large increase of major and life‐threatening bleeding complications in patients receiving DAPT as compared with patients receiving aspirin alone. Moreover, aspirin alone after TAVI was not associated with an increased incidence of the composite of thromboembolic complications (including all‐cause mortality, stroke, or myocardial infarction). Also, the individual incidences of all‐cause mortality, cardiovascular mortality, stroke, and myocardial infarction outcome were comparable between both groups.

Even decades after its introduction, the ideal antithrombotic strategy following TAVI is not completely determined yet. In order to minimize both the thromboembolic risk and the bleeding risk, an optimal balance in antithrombotic therapy should be pursued. The current guidelines on antithrombotic therapy following TAVI in patients without a long‐term indication for oral anticoagulation recommend DAPT with aspirin and clopidogrel for the first 3 to 6 months, followed by aspirin alone.⁷, ⁸ However, these statements rely on expert consensus, rather than prospective randomized evidence. Antiplatelet therapy in addition to aspirin is considered to reduce the risk of thromboembolic complications in the first period after TAVI, as was observed in patients undergoing coronary artery stenting.¹⁶ After this initial period, the thromboembolic risk is thought to return to that of the general elderly population and therefore does not need strong antithrombotic therapy.¹⁷

Obviously, initiation of dual antiplatelet therapy comes at the cost of an increased risk for bleeding, especially for patients undergoing TAVI. The general TAVI population is considered to be at higher risk for bleeding than patients undergoing coronary artery stenting, because of their advanced age and the presence of more comorbidities such as renal disease and previous bleeding. Previously randomized trials independently showed significantly lower incidences of bleeding events in patients receiving aspirin alone as compared with DAPT after TAVI.⁹, ¹⁰, ¹¹ In this meta‐analysis, we observed lower rates of major and life‐threatening bleeding in patients on aspirin alone compared with patients on DAPT.

More interesting is whether an aspirin‐alone strategy is effective enough in preventing thromboembolic complications as compared with DAPT. Of all thromboembolic complications, stroke is most prevalent, occurring in up to 8% within the first 30 days.³, ⁴, ⁵, ⁶, ¹⁸, ¹⁹, ²⁰ Studies investigating cerebral protection devices showed that a large amount of debris was scattered during the procedure, potentially causing stroke in the periprocedural period.²¹, ²² This debris contained acute thrombus, valve and artery wall tissue or calcification, and even foreign materials. In the subacute period (30 days to 3 months), other risk factors for stroke play a role such as existing or new‐onset atrial arrhythmias, incomplete stent frame endothelialization, and valve thrombosis (prosthetic or native valve thrombosis).²³, ²⁴, ²⁵ Based on these findings, the guidelines recommend DAPT in the early period after TAVI. On the other hand, several previously published observational and randomized studies observed no increase of thromboembolic complications, such as stroke, in patients receiving aspirin without additional antiplatelet therapy.⁹, ¹⁰, ¹¹, ²⁶, ²⁷, ²⁸, ²⁹, ³⁰, ³¹

Whether an aspirin‐alone strategy is sufficiently effective in preventing thromboembolic complications compared with DAPT is not known. All available trials were underpowered to answer this question. In this meta‐analysis we pooled all available randomized trials on a patient‐level, in order to amplify the statistical power and provide a broader insight in the efficacy of aspirin alone after TAVI. In the current study, we also observed no benefit of DAPT compared to aspirin alone after TAVI in preventing thromboembolic complications such as stroke at 30 days and at 3 months. In fact, the composite of thromboembolic complications and its individual components were comparable between both groups. None of the individual components showed a hint of an increased occurrence with aspirin alone. Based on our results, we consider aspirin alone the optimal antithrombotic treatment after TAVI rather than DAPT, in patients without a long‐term indication for oral anticoagulation and who do not require DAPT at the time of TAVI (eg, recent coronary artery stenting). On the other hand, the ongoing indication shift towards lower‐risk patients and a broader patient population undergoing TAVI might require a more patient‐tailored antithrombotic strategy, balancing the thrombotic and bleeding risk in the individual patient, instead of standardized treatments.

Our trial has several limitations. First, all included trials were open label and investigators were not blinded for treatment allocation, possible leading to a high risk for performance bias. Second, there was some heterogeneity regarding baseline characteristics between the included trials; however, the results of the included trials were all in the same direction. Third, the POPular TAVI trial was the largest trial and mainly responsible for the results of this meta‐analysis. On the other hand, the results of the smaller trials regarding both primary outcomes were in the same direction as in the POPular TAVI trial. Fourth, despite the fact that we included all available prospective randomized trials, this meta‐analysis remains relatively underpowered for the thromboembolic composite outcome. Last, the 3‐month follow‐up was only available for the POPular TAVI trial, ARTE trial, and Dual Antiplatelet Therapy Versus Aspirin Alone in Patients Undergoing Transcatheter Aortic Valve Implantation trial.

In this patient‐level meta‐analysis of randomized trials investigating aspirin alone as compared with aspirin with clopidogrel after TAVI in patients without a chronic indication for oral anticoagulation, aspirin alone was associated with a significant reduction of the composite of major and life‐threatening bleeding or thromboembolic events (including all‐cause mortality, major and life‐threatening bleeding, stroke, or myocardial infarction), and major or life‐threatening bleeding, compared with DAPT at 30 days and 3 months. Furthermore, aspirin alone, compared with DAPT, did not increase the incidence of the composite of thromboembolic events (all‐cause mortality, stroke, or myocardial infarction) at 30 days and 3 months.

Sources of Funding

None.

Disclosures

None.

Footnotes