Background

Dual antiplatelet therapy (DAPT) represents the pillar treatment for the management of patients with an acute coronary syndrome (ACS) and especially for those undergoing percutaneous coronary interventions, preventing periprocedural thrombotic complications and recurrent ischemic events [1,2,3].

More potent antiplatelet drugs and prolonged therapy [4,5,6], even beyond the traditional 12 months period, have certainly offered larger anti-ischemic benefits, and especially in certain higher-risk subsets of patients, as recently suggested in the PEGASUS-TIMI 54 trial [7]. However, an extended DAPT strategy has been associated with an excess of bleeding complications, and particularly in real life clinical settings, where the increased complexity of patients and the higher rate of comorbidities could increase both hemorrhagic and thrombotic risk [8, 9].

Advances in stent technologies, with the introduction of newer generations of drug eluting stents (DES), with thinner struts, bioresorbable polymer and faster re-endothelization have significantly reduced the rate of thrombotic complications, therefore allowing the discontinuation of DAPT after a shorter, 3–6 months period [10, 11].

However, despite the consensus of major experts and current evidence, that the indication for DAPT duration should be no longer 12 months in certain, low-risk subsets of stable patients undergoing PCI, the traditional 1-year strategy is still recommended in most of the patients after an acute coronary syndrome. Moreover, whether the optimal direction should be a shortening of DAPT, driven by the DES technology, or, on the contrary, a prolongation according to patient’s risk profile is still debated. In fact, few data had been reported so far specifically dedicated to the subset of patients undergoing PCI for an acute ischemic event, with very recently published trials providing comforting results on the shortening of DAPT in this setting, when patients are treated with newer generations of DES.

Therefore, the aim of the present meta-analysis was to assess the potential benefits of both a shorter-term (3–6 months) or prolonged (> 12 months) DAPT vs the standard traditionally recommended 12 months among ACS patients treated with PCI with DES.

Methods

Eligibility and search strategy

The literature was scanned by formal searches of electronic databases (MEDLINE, Cochrane and EMBASE) for clinical studies and the scientific session abstracts, searched on the TCT (www.tctmd.com), EuroPCR (www.europcr.com), ACC (www.acc.org), AHA (www.aha.org), and ESC (www.escardio.org) websites, for oral presentations and/or expert slide presentations from January 1990 to March 2019. The following key words were used: “dual antiplatelet therapy”, “duration”, “clopidogrel”, “prasugrel”, “ticagrelor”, “acute coronary syndrome”, “randomized”.

No language restrictions were enforced. Inclusion criteria were: (1) studies with randomized allocation to different duration of DAPT treatment; (2) availability of data specific to patients with acute coronary syndrome; (3) invasive management of patients with PCI; (4) availability of complete clinical and outcome data. Exclusion criteria were: (1) follow-up data in less than 90% of patients, (2) use of Bare Metal Stents > 10% of the patients; (3) ongoing studies or irretrievable data; (4) no allocation to a treatment arm comprising exactly 12 months DAPT therapy.

Data extraction and validity assessment

Data were independently abstracted by two investigators (MV, GDL). In case of incomplete or unclear data, the authors were contacted. Disagreements were resolved by consensus. Data were managed according to the intention-to-treat principle. Data on cardiovascular mortality were collected if data on overall mortality were not available.

Outcome measures

Primary efficacy endpoint was overall mortality. Primary safety endpoint was the rate of major bleeding complications (according to protocol definition).

Secondary endpoints were: (1) cardiovascular mortality, (2) recurrent myocardial infarction, and, (3) stent thrombosis (ST; definite or probable according to Academic Research Consortium-ARC definition); (4) Major cardiovascular ischemic events (MACE; per study definition).

Data analysis

Statistical analysis was performed using the Review Manager 5.3 freeware package, SPSS 23.0 statistical package. Odds ratio (OR) and 95% confidence intervals (95% CI) were used as summary statistics. The pooled odds ratio was calculated by using a fixed effect model (Mantel–Haensel). The Breslow-Day test was used to examine the statistical evidence of heterogeneity across the studies (p < 0.1). Potential publication bias was examined by constructing a “funnel plot”, in which sample size was plotted against odds ratios (for the primary endpoint). The study quality was evaluated by the same two investigators according to a score, that was expressed on ordinal scale, allocating 1 point for the presence of each of the following: (1) statement of objectives, (2) explicit inclusion and exclusion criteria, (3) description of intervention, (4) objective means of follow-up, (5) description of adverse events, (6) power analysis, (7) description of statistical methods, (8) multicenter design, 9) discussion of withdrawals, and, (10) details on medical therapy.

Adjusted indirect comparison of pooled estimates was then performed according to Biondi Zoccai et al. [12]. Specifically, we generated from fixed-effect OR, comparing shorter or extended DAPT vs 12 months therapy, an interaction OR for shorter (3–6 months) vs extended (> 12 months) DAPT, with pertinent 95% CI and Z scores for 2-tailed hypothesis testing (p significant if < 0.05).

We performed additional sensitivity analysis using Bayesian methods for pairwise meta-analysis with the bmeta R package (R Foundation for Statistical Computing, Vienna, Austria) and then repeated adjusted indirect comparisons according to Biondi-Zoccai et al. [12].

The study was performed in compliance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines [13].

Results

Eligible studies

A total of 298 studies were screened for inclusion in our meta-analysis, as shown in the flow-chart for the selection process in Fig. 1.

Fig. 1
figure 1

Flow diagram of the systematic overview process

Full size image

Finally, three studies [14,15,16] and six sub-analysis of randomized trials [17,18,19,20,21,22] were included in our meta-analysis, with an overall population of 15,738 patients. Among them, 8743 (55.5%) were randomized in trials comparing short-term vs standard DAPT strategy.

Two trials were excluded because of duplication [23, 24], 8 trials were excluded as data specific to patients with ACS were not provided [25,26,27,28,29,30,31,32,33,34]. Two trials [35, 36] were excluded because PCI was performed with BMS in a relevant proportion (> 10%) of the patients, while another trial [37] was excluded as DAPT duration did not include a treatment arm of exactly 12 months.

Antiplatelet agents used for the DAPT regimen were acetylsalicylic acid (ASA) and clopidogrel (75 mg/daily) in 4 trials, while in the other 5 trials new P2Y12 inhibitors (prasugrel, ticagrelor) were also used, or even preferred over clopidogrel [14, 16]. Use of glycoprotein IIbIIIa inhibitors was mainly not reported, although the majority of study protocols allowed it to the operators’ discretion, despite not encouraged.

Study characteristics of included trials are shown in Table 1. Mean follow-up was 16 ± 6.7 months, (median 12 months) ranging from 12 months [14,15,16,17,18,19,20,21] to 30 months [22].

Table 1 Characteristics of included randomized studies
Full size table

Table 2 displays the characteristics of enrolled patients, who had a mean age of 62.3 ± 2.2 years, included about 74.3% males, 25.5% of diabetics and 38.6% of patients with multivessel disease.

Table 2 Clinical features of patients in included studies
Full size table

Clinical outcome

Primary efficacy endpoint

Data on mortality were available in 15,705 patients (99.8% of total population). One trial provided only cardiovascular mortality [17]. A total of 276 patients (1.75%) had died at follow-up, with no significant difference in mortality with short-term or prolonged DAPT duration vs the traditional 12 months (short-term vs 12 months: 1.6% (71/4366) vs 1.6% (71/4377), OR [95% CI] 1.00 [0.72–1.39], p = 0.99; p for heterogeneity-phet = 0.41; prolonged vs 12 months: 1.8% (62/3473) vs 2% (72/3489), OR [95% CI] 0.87 [0.61–1.22], p = 0.41, phet = 0.92), as depicted in Fig. 2. Results did not change when restricting our analysis to the trials including only newer generation DES (short-term vs 12 months: OR [95% CI] 1.01[0.72–1.44], p = 0.94, phet = 0.23; prolonged vs 12 months: OR [95% CI] 0.86 [0.56–1.34], p = 0.51).

Fig. 2
figure 2

Benefits from short-term vs standard DAPT (dual antiplatelet therapy) (upper part) and from prolonged vs standard DAPT on mortality with odds ratios and 95% confidence intervals (CI). The size of the data markers (squares) is approximately proportional to the statistical weight of each trial. *Defined as cardiovascular mortality

Full size image

Primary safety endpoint

Data on major bleedings were available in 15,705 patients (99.8%). A major bleeding event was documented in 210 patients (1.3%). Thrombolysis in myocardial infarction (TIMI) Major definition was used in four studies [15, 17, 18, 21], whereas three studies provided data on BARC (Bleeding Academic Research Consortium definition) [14, 16, 19].

As displayed in Fig. 3, the shorter DAPT regimen offered a non- significant reduction in major bleeding events (0.9% (42/4366) vs 1.1% (48/4377), OR [95% CI] 0.87 [0.58–1.33], p = 0.53, phet = 0.41) while the prolongation of DAPT beyond 12 months was associated to an increased risk of hemorrhagic complications (2.2% (75/3473) vs 1.3% (45/3489); OR [95% CI] 1.69 [1.17–2.45], p = 0.006; phet = 0.23, NNH = 111). Results did not change when restricting our analysis to the trials including only newer generation DES (short-term vs 12 months: OR [95% CI] 0.93 [0.60–1.44], p = 0.74, phet = 0.22; prolonged vs 12 months: OR [95% CI] 5.42 [0.26–113.80], p = 0.28), although the latter being represented only by the ARCTIC Interruption trial [20]. Similar findings were observed when using only the BARC definition for major bleedings (short-term vs 12 months: OR [95% CI] 0.93 [0.60–1.44], p = 0.74, phet = 0.22; prolonged vs 12 months: not applicable).

Fig. 3
figure 3

Shorter versus longer DAPT (dual antiplatelet therapy) duration on major bleedings with odds ratios and 95% confidence intervals (CI). The size of the data markers (squares) is approximately proportional to the statistical weight of each trial

Full size image

Secondary endpoints

Cardiovascular mortality

Data on cardiovascular mortality were available in 14,614 patients (92.9%); among them, 143 (0.98%) experienced an event, with no difference with respect to DAPT duration strategy (as in Fig. 4). Similar results were obtained in trials on short-term vs traditional DAPT (0.97% (35/3752) vs 1.1% (38/3570), OR [95% CI] 0.92 [0.58–1.46], p = 0.72, phet = 0.41), and with prolonged vs 12 months DAPT (0.92% (32/3473) vs 0.95% (38/3999), OR [95% CI] 0.85 [0.53–1.36] p = 0.50, phet = 0.60).

Fig. 4
figure 4

Shorter versus longer DAPT (dual antiplatelet therapy) duration on cardiovascular mortality with odds ratios and 95% confidence intervals (CI). The size of the data markers (squares) is approximately proportional to the statistical weight of each trial

Full size image

Myocardial infarction

Data on recurrent myocardial infarction were available in 15,705 patients (99.8%), of whom 338 (2.2%) experienced an event. The risk of non-fatal myocardial infarction was comparable with short-term vs 12 months standard DAPT strategy (2% (89/4366) vs 1.6% (72/4377), OR [95% CI] 1.24 [0.91- 1.70], p = 0.18, phet = 0.34), while prolonged DAPT duration (> 12 months) could provide a significant reduction in the rate of events (1.7% (59/3473) vs 3.3% (118/3489); OR [95% CI] 0.49 [0.36–0.67], p < 0.00,001; phet = 0.38; NNT = 62.5), as shown in Fig. 5.

Fig. 5
figure 5

Shorter versus longer DAPT (dual antiplatelet therapy) duration on myocardial infarction with odds ratios and 95% confidence intervals (CI). The size of the data markers (squares) is approximately proportional to the statistical weight of each trial

Full size image

Stent thrombosis (ST)

Data on definite/probable ST were available in 15,705 patients (99.8%). A total of 114 (0.7%) experienced such an event. The risk of ST was not significantly increased by shortening DAPT duration (0.76% (33/4366) vs 0.48% (21/4377), OR [95% CI] 1.57 [0.91–2.70], p = 0.11; phet = 0.81), while significant benefits in ST were observed with an extended DAPT duration (0.49% (17/3473) vs 1.2% (43/3489) OR [95% CI] 0.40 [0.23–0.70], p = 0.0001; phet = 0.19; NNT = 141), as in Fig. 6.

Fig. 6
figure 6

Shorter versus longer DAPT (dual antiplatelet therapy) duration on stent thrombosis (definite/probable) with odds ratios and 95% confidence intervals (CI). The size of the data markers (squares) is approximately proportional to the statistical weight of each trial

Full size image

Major cardiovascular ischemic events (MACE)

Data on MACE were available in 12,129 (77.1%) of the patients. A total of 544 (4.5%) experienced such an event. The risk of MACE was not significantly increased by innovative DAPT strategies as compared to the traditional 12 months (4.6% (281/6084) vs 4.3% (263/6045), OR [95% CI] 1.07 [0.90–1.27], p = 0.47; phet = 0.75), with similar results for a shorter or a longer DAPT duration (respectively OR [95% CI] 1.10 [0.91–1.33], p = 0.32; phet = 0.66 and OR [95% CI] 0.90 [0.59–1.38], p = 0.64; phet = 0.60), Fig. 7.

Fig. 7
figure 7

Shorter versus longer DAPT (dual antiplatelet therapy) duration on major cardiovascular events (MACE) with odds ratios and 95% confidence intervals (CI). The size of the data markers (squares) is approximately proportional to the statistical weight of each trial

Full size image

Adjusted indirect comparison

Indirect comparison of 3–6 months vs extended (> 12 months) DAPT alone showed a significantly increased protection against recurrent myocardial infarction and stent thrombosis when prolonging DAPT beyond 12 months (p < 0.001 and p = 0.0006 respectively), although this strategy was weighted by a significant increase in major bleeding complications (p = 0.02) (Table 3 and Fig. 8).

Table 3 Indirect comparison between 3–6 vs > 12 months DAPT
Full size table
Fig. 8
figure 8

Indirect comparison for shorter (3-6) versus > 12 months DAPT (dual antiplatelet therapy) duration with odds ratios and 95% confidence intervals (CI)

Full size image

Sensitivity analysis using Bayesian methods for pairwise and adjusted indirect comparisons confirmed the main results stemming from a frequentist framework in terms of direction and strength of effect (Table 4).

Table 4 Bayesian pairwise meta-analysis and ensuing adjusted indirect comparison focusing on different durations of dual antiplatelet therapy
Full size table

Discussion

The present study represents the most comprehensive and updated meta-analysis addressing the optimal duration of DAPT in patients with ACS undergoing percutaneous coronary revascularization with new generation drug-eluting stents.

We demonstrated that a short-term 3–6 months DAPT strategy can offer a comparable protection from major cardiovascular ischemic events than the standard 12 months strategy currently indicated in guidelines. Extending the duration of the therapy beyond 12 months could further empower the antithrombotic protection, although paying the fee of increasing major bleeding complications, therefore not resulting in any survival benefit.

The impressive growth of interventional cardiology and the significant increase in the number of patients with acute coronary syndromes undergoing early percutaneous coronary revascularization have led in the last years to an escalation in the complexity of the management of DAPT, and particularly for its optimal duration [38,39,40].

In fact, whereas in stable patients, a tendency towards a progressive shortening of the period of antiplatelet treatment has gone along with the technological development of newer, less thrombogenic, drug-luting stents (DES) [11, 41], a similar approach could not fit to the higher-risk patients admitted for an acute cardiovascular event [42, 43], where both hemorrhagic risk and cardiovascular events are enhanced [44,45,46].

Therefore, the field of interventional cardiologists has been required to deal with these two opposite trends, with DES technologies performing even better than bare metal-stents and reducing the rate of thrombosis to neglectable levels (< 1%) [41], therefore allowing to shorten the duration of DAPT and preventing potential hemorrhagic complications; while oppositely high-cardiovascular risk patients’ profile was associated to larger benefits from prolonging DAPT even beyond the traditional 12 months [7].

Several guidelines and consensus have attempted in the past years to shed light on the optimal timing for DAPT discontinuation in ACS patients [2, 47], although not being supported by dedicated randomized trials, but only from data derived from sub-analysis of studies and meta-analysis, conducted with older generations of DES. The indication to 12 months DAPT itself, in effect, derives from an old trial, the PCI-CURE [35], where the relevant quote of the benefits was achieved very early after the randomization and before angiography. In fact, patients randomized to 12-month DAPT in the PCI-CURE trial started clopidogrel soon after randomization whereas those patients randomized to 1-month DAPT received placebo before angiography. The landmark analysis clearly showed that a relevant quote of benefits was observed before angiography and that the difference between both study groups became non-significant after 90 days post PCI. In the light of this observation, the optimal duration of DAPT in ACS patients is still subject of controversy.

Recently, three randomized trials have been conducted, specifically designed to assess the safety of the interruption of DAPT before 12 months in ACS patients treated with newer generations of DES [14,15,16]. Results were concordant in reporting the non-inferiority of a shorter 3–6 months DAPT regimen as compared to the standard treatment in these patients, therefore arising again the issue of shifting the indication of the optimal DAPT duration to less than 1 year.

However, the relatively small number of patients and events in these studies, not completely powered for the evaluation of secondary endpoints such as MI or ST, and the presence of a still modest increase in the rate of thrombotic complications with a 3–6 months DAPT, still raises some uncertainty on changing the actual indications, at least for low-risk ACS patients.

The present meta-analysis represents the most updated study on the optimal DAPT duration in the settings of ACS patients treated with PCI and new generation DES. We found a similar outcome between a short 3–6 months vs the traditional 12 months DAPT regimen for any of the study endpoints. This finding may certainly question current recommendation of routine 12 months DAPT. In fact, even though a non-significant trend towards increased ischemic events was observed with a short DAPT, on the contrary, as shown by the indirect comparison analysis, to foresee a further reduction of recurrent cardiovascular events, DAPT should have been extended to more than 12 months, although at the expense of increasing bleedings. Sensitivity analysis using Bayesian methods for pairwise and adjusted indirect comparisons confirmed the main results stemming from a frequentist framework in terms of direction and strength of effect.

However, it must be admitted that results on MI and ST in the subgroup of prolonged DAPT were mainly driven by a sub-analysis of the DES LATE trial [21] and DAPT study [22], performed with older DES and with mainly with clopidogrel, whereas no dedicated trial has ever been performed with DAPT beyond 1 year, specific for the ACS population and with the use of more potent antithrombotic agents currently recommended in guidelines, such as ticagrelor or prasugrel.

Similar results to ours were recently reported in a network meta-analysis by Palmerini et al. [48], where 3 but not 6 months DAPT significantly increased MI or stent thrombosis in ACS patients. Moreover, a shorter DAPT duration was associated with a lower rate of hemorrhagic complications, irrespective of presentation, whereas mortality was not conditioned by the timing of DAPT discontinuation. However, this meta-analysis was conducted only on dated trials, with still a large use of BMS and older DES, moreover no study specific for the ACS population was included, and a strategy of extended antithrombotic therapy beyond 1 year was not evaluated.

On the contrary, the issue of prolonging DAPT beyond the traditional 12 months, in higher risk ACS patients, was addressed in another recent meta-analysis by Bavishi et al. [49], driven by the positive results of the study of Udell et al. [50] and of the Prevention of Cardiovascular Events in Patients with Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin Thrombolysis in Myocardial Infarction 54 (PEGASUS-TIMI 54) trial, (7) although these studies were not specifically designed to investigate the different duration of DAPT post-stenting. Analogous conclusion can be reached from present data.

Keeping into account the costs of prolonged DAPT, especially in the era of new ADP antagonists, a potential routine strategy of routine 3–6 months DAPT may be considered, leaving a longer DAPT in those patients judged at high-risk for thrombotic complications if this exceeds the risk of major bleeding. Indeed, future larger studies are certainly needed in this subset of patients, since the insufficient power of the present meta-analysis certainly conditions the generability of our results.

Study limitations

The first limitation to our study derives from the synthesis of different trials, with no access to individual patients’ data. Therefore, we were not able to provide any data of the magnitude of ACS risk or to provide a subgroup analysis according to ACS risk scores. Despite, no significant heterogeneity was found in our endpoints, the definitions of major bleeding differed among trials. However, two definitions (TIMI and BARC) were used in the majority of trials, representing the most largely applied definitions for major bleedings, with a validated impact on the outcomes in cardiovascular trials and especially in the subset of ACS [51].

Nevertheless, we were not able to provide data on the net clinical benefit endpoints. In fact, we preferred to evaluate the individual events in order to prevent the heterogeneity due to the use of different definitions for the composite endpoints in the different studies. Neither we could evaluate the occurrence of cerebrovascular events among our study endpoints since data were available only for the overall study population and not specific to the ACS subset of patients, in particular for trials comparing extended vs 12 months DAPT [52]. In effect, such limitation was previously experienced by Bavishi et al. [49], that, despite including stroke among secondary endpoints, did not provide any pooled results for such endpoint.

In addition, the combination of DAPT drugs and PCI strategy were largely different among the trials, accounting for almost 10-years of development of percutaneous coronary revascularization techniques (from the design of older trials-DES LATE, 24 in 2010- to present).

Moreover, even if we excluded the trials allowing the use of BMS, still the majority of patients underwent PCI with older generations of DES, therefore not being comparable to the current standard of care in terms of PCI. In fact, the most positive effects with shorter DAPT were observed in the most recent trials. Unfortunately, we were forced to exclude some of the most recent trials with newer generations of DES due to the unavailability of specific data on the ACS population. Nevertheless, we provided a separate analysis restricted to the use of new generation DES and accounting for the most updated definition of bleedings (BARC), showing no difference between a shorter vs 12 months DAPT. However, no conclusion could be drawn for prolonged DAPT strategies.

In addition, it might be expected that the increase of bleedings with a prolonged DAPT strategy would have been even more marked it the antiplatelet regiment would have been composed of other drugs instead of clopidogrel, thus further reinforcing the hesitation in extending DAPT.

Finally, it might be argued that the modest number of events, especially in most recent trials, could have prevented the observation of any statistical difference between the two treatment arms, should there be any. However, the extremely low rate of ischemic events, together with the null impact on mortality, further reinforce the need of re-considering the indications about DAPT, in order to progressively shortening its optimal duration. Nevertheless, despite representing the most comprehensive study on the ACS population, still our study was underpowered to detect significant differences in mortality. In fact, based on the observed 1.6% and 2% mortality rate in the 12 months DAPT arm, and based on a superiority design, with a 2-sided test for differences in independent binomial proportions at the 5% significance level, a power of 80%, a total of 16,722 and 14,572 patients, respectively, would have been required to show a 30% relative risk reduction of mortality with shorter and prolonged vs standard 12 months DAPT strategy.

Conclusion

Based on the current meta-analysis, among ACS patients treated with percutaneous coronary interventions with DES, a shorter 3–6 months DAPT can be safely considered, offering a non-inferior protection from major cardiovascular ischemic events as compared to the standard 12 months strategy. Extending DAPT therapy beyond 12 months enhances the antithrombotic protection, although paying the fee of increasing major bleeding complications, therefore resulting in a null effect on mortality.