Figure 2 shows the timeline of key randomized clinical trials of antiplatelet therapy after CABG.

Fig. 2

Randomized clinical trials of antiplatelet therapy after CABG. The figure presents a visual representation of randomized clinical trials investigating antiplatelet therapy following CABG

Single antiplatelet therapy

Early placebo-controlled studies published in the eighties tested warfarin, indobufen, aspirin at different doses and the combination of high-dose aspirin and dipyridamole, with mixed results [14,15,16,17,18,19,20,21,22,23]. In the largest of these studies (n = 555) all the aspirin-based regimens improved graft patency at 60 days from surgery [19]. These results were consistent in a smaller trial (n = 231) where an aspirin dose of 324 mg daily, given within one hour after CABG, resulted in a significant reduction in SVG occlusion at 1 week that was sustained at one year [18]. Following these studies, aspirin became a pillar of secondary prevention in this setting [24, 25]. However, the response to aspirin can be highly variable among CABG patients [26, 27]. Therefore, alternative antiplatelet agents, including the P2Y12 receptor inhibitors clopidogrel and ticagrelor, have been investigated (Table 1).

Table 1 Randomized trials comparing monotherapy regimens after CABG Clopidogrel

Clopidogrel irreversibly inhibits the platelet response through mechanisms mediated by adenosine diphosphate. A subgroup analysis of patients undergoing cardiac surgery in the CAPRIE trial (including but not limited to patients undergoing CABG) suggested that clopidogrel might be better than aspirin in the context of long-term management [28, 29]. On the other hand, several pharmacodynamic studies failed to demonstrate a relevant early effect of clopidogrel in comparison with other single antiplatelet agents. For example, a study of 62 patients compared different doses of clopidogrel (i.e., 50, 75, or 100 mg) with ticlopidine 250 mg given twice daily, and found that all the three doses effectively inhibited platelet activity ex-vivo and prolonged bleeding time at day 28, but did not significantly reduce platelet aggregation at day 9 [30]. Another small study of 54 patients compared the effect of two doses of aspirin (100 mg or 325 mg) with clopidogrel 75 mg, and found no significant benefit of clopidogrel on platelet aggregation during the first five postoperative days [31]. As such, clopidogrel monotherapy is not currently recommended over aspirin to prevent the risk of early graft failure.

Ticagrelor

Ticagrelor reversibly binds to the platelet P2Y12 receptor and provides strong and rapid inhibition of adenosine diphosphate -induced platelet aggregation. The TiCAB trial compared ticagrelor monotherapy (90 mg, twice daily) with aspirin monotherapy (100 mg/day) during the first year after arterial and/or SVG implantation [32]. The trial was discontinued prematurely due to withdrawal of funding support from the sponsor, at a time when 1859 out of 3850 planned patients were randomized. No significant differences were observed between ticagrelor and aspirin in terms of major adverse cardiac events (MACE) at 12 months after CABG. However, these results should be interpreted with caution as the study was underpowered. The TARGET trial was another relatively small study (n = 250) comparing ticagrelor and aspirin after CABG [33]. The primary outcome was occlusion of SVG as determined by computed tomography coronary angiography at 12 months. There was no significant difference between the two groups [33], which was also confirmed at two years in 142 patients undergoing another computed tomography coronary angiography assessment [34]. In aggregate, similarly to clopidogrel, there is no evidence to recommend ticagrelor as single antiplatelet therapy after CABG.

Dual antiplatelet therapy

Randomized clinical trials of DAPT for the prevention of graft occlusion are summarized below and in Table 2.

Table 2 Randomized clinical trials of dual versus single antiplatelet strategies after CABG DAPT with aspirin and clopidogrel

Several studies compared the combination of clopidogrel and aspirin versus aspirin monotherapy, with small study samples and mixed findings. Four studies have reported improved graft patency [35,36,37,38]. The largest of these studies was the CRYSSA trial (n = 300), which showed a significantly lower risk of graft occlusion at 12 months [37]. Conversely, four other studies presented negative findings [39,40,41,42].

No trials of DAPT with clopidogrel and aspirin exist that were powered for hard clinical endpoints. Data from observational studies and post-hoc analyses of randomized trials designed for other purposes are available, but likely fraught by confounding bias. If anything, these studies were consistent in showing no benefit on mortality with clopidogrel in addition to aspirin. In the 2,072 patients with acute coronary syndromes (ACS) who received CABG in the CURE trial, DAPT reduced the risk of cardiovascular death, myocardial infarction, or stroke by 11%, but increased the risk of hemorrhagic complications by 30% [43]. Conversely, in a large retrospective registry from China (n = 18,069), CABG patients with DAPT had a lower incidence of all-cause death, stroke, myocardial infarction, or repeat revascularization at six months and had no differences in bleeding events [44]. In a sub-analysis of the ROOBY trial, DAPT increased early death (i.e., within 30 days) and did not improve the risk of death at long-term [45]. Additionally, in a study including aspirin-resistant patients, DAPT did not result in reduced death at six months [46]. These results were consistent with a post hoc analysis of the FREEDOM trial, where no differences in death, adverse ischemic events and bleeding was observed at five years among patients with type II diabetes mellitus undergoing CABG [47].

DAPT with aspirin and ticagrelor

The TAP-CABG trial (n = 70), which was terminated prematurely because of slow recruitment, evaluated the incidence of arterial and venous graft patency at 3 months after DAPT with ticagrelor and aspirin versus aspirin alone. The primary endpoint was slightly improved with DAPT (p = 0.044), but the difference was not significant in analyses stratified by individual grafts [48]. The larger DACAB trial randomized 500 patients to DAPT, ticagrelor alone or aspirin alone [49]. DAPT significantly improved the rate of SVG patency at 12 months compared to aspirin (risk difference, 12.2; 95% confidence interval [CI], 5.2–19.2%; p < 0.001). This effect was consistent in a post hoc analysis restricted to patients with ACS, who represented 67% of the entire population. The incidence of ischemic and bleeding events was low, which precludes the interpretation of clinical endpoints. A 5-year follow-up extension study of DACAB, where more events will be accrued, is ongoing (NCT03987373).

At variance with DACAB, the POPULAR-CABG trial (n = 499) showed no significant difference in one-year SVG patency with aspirin and ticagrelor compared to aspirin alone [50]. The different results of DACAB and POPULAR-CABG have two contributing explanations. Firstly, in DACAB, a higher proportion of patients underwent CABG for ACS than in POPULAR-CABG (i.e., two thirds versus one third). ACS is known as the population that benefits the most from a ticagrelor-based DAPT. Secondly, the use of cardiopulmonary bypass was markedly lower in DACAB (25%) than in POPULAR-CABG (95%). The impact of this difference on graft patency is unclear. Two recent meta-analyses suggested that the antiplatelet regimens that include ticagrelor are associated with improved clinical outcomes and increased graft patency [51, 52]. However, the findings of these meta-analyses were mixed regarding the risk of clinically important bleeding. In view of the conflicting results of the available studies, the efficacy of DAPT with ticagrelor and aspirin in improving the patency of SVGs remains undefined.

DAPT with aspirin and prasugrel

A recent study compared DAPT with prasugrel and aspirin versus aspirin alone [53], but was prematurely stopped due to slow enrolment after randomizing only 84 patients. The primary endpoint, the incidence of optical coherence tomography-detected SVG thrombus at 12 months, was observed in approximately one-third of the patients, without a significant difference between the two treatment groups. Additionally, there were no significant differences in angiographic SVG failure, the incidence of MACE, or severe bleeding. Two meta-analyses suggested that DAPT with prasugrel reduces the risk of SVG failure, mortality and MACE when compared with single antiplatelet therapy, albeit at the expense of an increased risk of major bleeding [54, 55].

Comparisons of DAPT strategies

Although the evidence in this area is not robust, some post-hoc analyses of studies comparing different DAPT strategies are informative. DAPT with aspirin and clopidogrel was compared to DAPT with aspirin and ticagrelor in the PLATO trial, which included 1,261 patients with ACS undergoing CABG [56]. The results in this subgroup showed a significant reduction in the composite outcomes of all-cause death, myocardial infarction, or stroke at 12 months with aspirin and ticagrelor, and similar rates of hemorrhagic events. Additionally, a pharmacodynamic study conducted in 140 patients undergoing CABG demonstrated that the onset of action was faster and the inhibition of platelet aggregation was higher with ticagrelor and aspirin than with clopidogrel and aspirin, with no difference in bleeding or MACE [57]. Another small trial (n = 147) reported similar rates of SVG patency at 1-year with ticagrelor-based and clopidogrel-based DAPT [58].

The only available data comparing DAPT with prasugrel and aspirin and DAPT with clopidogrel and aspirin comes from a subset analysis of the TRITON-TIMI 38 trial, which included 346 patients with ACS undergoing CABG [59]. Despite an increase in bleeding and surgical re-exploration, prasugrel-based DAPT was associated with a lower rate of death within 30 days after CABG. It is possible that the greater degree of platelet inhibition provided by prasugrel may have contributed to both the increased non-fatal bleeding and the reduced risk of fatal cardiac events and mortality. This evidence is mostly derived from sub-analyses of trials with non-stratified randomization, and therefore is not sufficient to draw definitive conclusions.

A recent Chinese trial (n = 152) compared DAPT with indobufen and clopidogrel to DAPT with aspirin and clopidogrel and found similar patency rates of SVG and arterial grafts at 12 months [60]. This trial also showed a similar rate of MACE between the two groups and a lower incidence of gastrointestinal adverse events in the indobufen group. Based on these findings, indobufen might be considered in DAPT combinations if aspirin is not an option.

Anticoagulant therapy

Early studies investigated the effectiveness of various anticoagulants in preventing graft occlusion after CABG.

Vitamin K antagonists

In 1993, a meta-analysis of 17 trials concluded that warfarin significantly reduces the risk of graft occlusion compared to placebo, similar to aspirin [61]. No difference between vitamin K antagonists (VKA; i.e., acenocoumarol or phenprocoumon) and aspirin was demonstrated on SVG patency at one year in a trial of 948 patients [62].

In the landmark Post-CABG (Post-Coronary Artery Bypass Graft) trial, 1,351 patients on aspirin were randomized to low-dose warfarin (e.g., dual-pathway inhibition) or placebo [63]. While no significant effect was observed on progression of SVG disease, there were a 35% reduction in mortality (p = 0.008) and a 31% reduction of death or nonfatal myocardial infarction (p = 0.003) with warfarin and aspirin at 7.5 years [64]. The mechanism leading to such effects remained unexplained and play of chance cannot be ruled out. Indeed, only 11% of patients were on VKA during the extended follow-up.

Direct oral anticoagulant

More recently, there has been interest in evaluating a strategy of combining a direct oral anticoagulant (DOAC) with an antiplatelet agent. The rationale for this strategy is to reduce the degree of platelet activation throughout synergistic inhibition of thromboxane A2 production by aspirin and inhibition of thrombin and fibrin formation by the DOAC [65]. Due to lack of dedicated trials, whether this strategy is suitable for secondary prevention after CABG is unclear [66, 67].

In a prespecified substudy of the COMPASS trial, 1,448 patients were randomized within 4 to 14 days after CABG to rivaroxaban 2.5 mg twice daily plus aspirin 100 mg daily, rivaroxaban 5 mg twice daily, or aspirin 100 mg daily [68]. At an average of 1.13 years, compared to aspirin alone, rivaroxaban did not reduce the rate of both arterial and SVG failure either as a combination with aspirin or as monotherapy. Additionally, the two rivaroxaban-based strategies did not reduce the risk of a composite of cardiovascular death, stroke, or myocardial infarction and increased the risk of bleeding at 30 days after CABG. Notably, when compared to aspirin alone, the combination of rivaroxaban and aspirin did not increase the rate of graft patency in both patients treated with on-pump and off-pump techniques. Conversely, rivaroxaban monotherapy improved the rate of graft patency in patients undergoing off-pump CABG (odds ratio 0.37; 95% CI, 0.16 to 0.82; p = 0.01), but not in those undergoing on-pump CABG. Overall, these results do not support the use of rivaroxaban, either alone or in a dual-pathway inhibition regimen, after CABG. Further studies are warranted to corroborate the promise of rivaroxaban in patients undergoing off-pump CABG.

Parenteral anticoagulation

Another approach to prevent early graft failure is the use of parenteral anticoagulants such as fondaparinux. In the Fonda-CABG study, 99 CABG patients on aspirin therapy were randomized to fondaparinux 2.5 mg/daily or heparin in the early postoperative in-hospital period [69]. After discharge and up to 30 days, the fondaparinux group continued to receive fondaparinux, while the heparin group received placebo. Computed tomography angiography performed at 30 days demonstrated similar rates of graft occlusion and no statistically significant difference in death, stroke, myocardial infarction, bleeding events, or re-operation. Although it was not adequately powered for efficacy, the trial showed no benefit of extended fondaparinux therapy compared with heparin for the prevention of early graft failure.