Currently, the standard antithrombotic therapy for secondary cardiovascular prevention after an acute coronary syndrome (ACS) consists of dual antiplatelet therapy (DAPT) for up to 12 months.1–3 However, there is still much uncertainty regarding the ideal duration of antithrombotic therapy, particularly when considering patients’ different ischemic and bleeding risk profiles.4–6 Beyond 12 months, there is still a considerable risk of ischemic events, and a small number of studies have examined the outcomes of extended DAPT.7–11 This led to the publication of a focused update on DAPT by the European Society of Cardiology and the European Association of Cardio-Thoracic Surgery.12

Until recently, the role of the new oral anticoagulants and ticagrelor in long-term regimens in patients with ACS was uncertain. However, these drugs were assessed in two recent randomized clinical trials (RCTs) that set out to improve on the existing standard of care for secondary cardiovascular prevention after an ACS. Both trials indicated significant improvements regarding cardiovascular outcomes and death. However, the large number of eligibility criteria used in these RCTs may limit the applicability of their results in clinical practice.

The purpose of our study was to assess the eligibility for long-term use of low-dose (LD) ticagrelor (60mg twice daily) or LD rivaroxaban (2.5mg twice daily) according to the enrollment criteria of two trials of extended antithrombotic treatment after ACS, Long-Term Use of Ticagrelor in Patients with Prior Myocardial Infarction (PEGASUS-TIMI 54)13 and Rivaroxaban with or without Aspirin in Stable Cardiovascular Disease (COMPASS),14 respectively, in a cohort of consecutive ACS patients.

A total of 1021 patients were admitted for ACS during the 18-month inclusion period. Of those, 938 underwent PCI. Median length of stay was 6.0 days (IQR 3.0-11.0). Regarding the type of ACS, 30.3% were classified as STEMI, 54.4% as NSTEMI and 15.3% as UA. In-hospital mortality was 7.9%. Mortality within 12 months of discharge was 9.7%.

A total of 780 patients were screened for enrollment in the RCTs after 12 months of follow-up (Figure 1). Baseline characteristics are summarized in Table 2. The proportion of patients fulfilling the trial enrollment criteria was not significantly different between the trials (35.9%, n=250 for PEGASUS vs. 32.1%, n=280 for COMPASS, p=0.113) (Figure 2). The proportion of patients eligible for both RCTs was 17.7% (n=138) and 49.7% (n=388) of patients were eligible for at least one of them (Figure 3). The main reasons for exclusion from each RCT were the need for oral anticoagulation (OAC) in PEGASUS (46.2%, n=231) and the presence of high bleeding risk in COMPASS (61.5%, n=326). The need for OAC was also an important reason for exclusion in COMPASS (32.1%, n=170) and occurred in 22.4% (n=210) of the total study population. The main reasons for OAC are summarized in Table 3. Other significant reasons for exclusion are illustrated in Figure 4. Patients not fulfilling enrollment criteria for either RCT (58.4%) were older (71.4±9.7 vs. 65.2±10.2 years, p<0.0001) and had worse renal function (creatinine clearance 47.8±28.8 vs. 73.9±21.2ml/min/1.73 m2, p<0.0001), lower left ventricular ejection fraction (LVEF) (44.4±11.4 vs. 51.9±7.5%, p<0.0001) and longer hospital stay (7 [IQR 4-13] vs. 5 days [IQR 3-8], p<0.0001). The main differences between patients eligible and ineligible for both RCTs are summarized in Table 4. Compared with non-eligible patients (n=388), patients who were eligible for both RCTs (n=138) were typically younger (64.4±8.7 vs. 70.3±10.1 years, p<0.001) and had higher LVEF (53.8±5.1 vs. 44.6±11.5%, p<0.001), higher creatinine clearance (73.6±18.2 vs. 47.7±29.1ml/min/1.73 m2), and lower prevalence of cardiovascular risk factors (Table 5).

Figure 1.

Patient flow-chart. ACS: acute coronary syndrome; CABG: coronary artery bypass grafting; PCI: percutaneous coronary intervention.

(0,29MB).

Figure 2.

Distribution of patients according to the eligibility criteria in the two randomized controlled trials.

(0,1MB).

Figure 3.

Distribution of patients eligible for the two randomized controlled trials.

(0,06MB).

Figure 4.

Reasons for exclusion from the two randomized controlled trials. CKD: chronic kidney disease; DAPT: dual antiplatelet therapy; NYHA: New York Heart Association class; LVEF: left ventricular ejection fraction; OAC: oral anticoagulation.

(0,18MB).

Our results show that, when the eligibility criteria of the PEGASUS-TIMI 54 and COMPASS trials are applied to a real-world population of consecutive patients with previous ACS, only around one third are eligible for the novel long-term antithrombotic therapies. The authors suggest that as soon as these drugs are available in Portugal, prospective registries should be established to assess their overall benefit in patients not eligible for the two RCTs under study, as has been done with the new oral anticoagulants.

It can be difficult to perform external validation and to determine the applicability of RCTs, since the strict enrollment criteria they use often result in a highly selected population that is usually at lower risk than patients in everyday practice, and who may not be representative of real-world populations regarding characteristics, risk factors and outcomes.15–18

It is known that more intense antithrombotic therapy reduces ischemic complications, but this comes at the cost of more bleeding events.13,19,20 Trial designers therefore often aim to exclude patients with high bleeding risk. However, our results show that real-world patients with previous ACS have higher risk scores for both cardiovascular and bleeding events than those enrolled in the PEGASUS and COMPASS trials.

As previously reported, patients who are excluded from RCTs usually represent high-risk subgroups with poor outcomes.21,22 In our study, patients who met the exclusion criteria for both the RCTs under study were those with the highest ischemic risk, as indicated by the significant differences in GRACE scores, and hence those who would most likely benefit from extended antithrombotic therapy. However, compared with eligible patients, they were older and had higher bleeding risk scores and more severely impaired renal function. This paradoxical relation between bleeding and thrombotic risk is reported in the literature.23,24

The need for chronic anticoagulation was an important reason for exclusion, found in 22.4% of the total study population (46.2% of excluded patients on the PEGASUS criteria and 32.1% of excluded patients on the COMPASS criteria). The exclusion of patients due to high bleeding risk is in line with other registries, however the proportion of patients ineligible for the COMPASS trial due to bleeding risk in our population was higher (28.9% in the REACH registry and 17% in the FAST-MI registry).25,26 This could be explained by the addition of bleeding score risks in our population, since the bleeding risk assumption in the COMPASS trial was left to the investigators’ discretion.14 Nevertheless, in both trials, patients had to undergo at least 12 months of DAPT without bleeding events to be eligible. This may have selected a specific population with lower bleeding risk.

In our study, we applied the enrollment criteria to patients 12 months after an ACS. This population also does not correspond to that reported in the COMPASS trial, inasmuch as although 70% of the patients enrolled had a previous myocardial infarction, around half of these had occurred more than five years prior to randomization, which may have altered their long-term ischemic outcomes compared with a real-world population. On the other hand, more than half of ACS patients in PEGASUS had STEMI and were at high risk for ischemic events. Hence, patients at low risk for ischemic events may not benefit as much as seen in the trial.

This real-world cohort registry included patients with similar characteristics to those reported in other published studies on the maintenance of prolonged DAPT27,28 that reported no significant differences in mortality, despite reductions in ischemic events, due to significant increases in bleeding events.

The patients eligible for trial enrollment were similar across the RCTs, suggesting that the deciding factor in choosing between LD ticagrelor or LD rivaroxaban is not specific patient characteristics, but mainly the time elapsed since the last ACS.

To our knowledge, this is the first study to apply the enrollment criteria of the extended antithrombotic therapy trials to a Portuguese population, and the first anywhere to apply the criteria from both trials in the same population. The strength of the study lies in the inclusion of a reasonable number of unselected and consecutively enrolled patients. However, the limitations resulting from a single-center observation should also be considered. Further studies enrolling high-risk patients are needed to clarify the benefit of these interventions in more representative populations. As these therapies are still not available in Portugal, the authors suggest the development of prospective registries after these drugs are approved and enter the market, to assess their overall benefit in patients not eligible for the RCTs under study, as has been done with the new oral anticoagulants.29

Patients eligible for trial enrollment in PEGASUS and COMPASS are only partly representative of the real population with a recent history of ACS. Hence, their results are limited if applied to the real-world patients that are at greater risk of future events, who represent approximately half of the total population. Further studies with less restrictive enrollment criteria are needed to determine if this approach is still beneficial in these patients.

Daniel Faria has received speaker fees from AstraZeneca, Miguel Santos has received speaker fees from AstraZeneca and Bayer, and Sérgio Baptista has received research grants from AstraZeneca and Bayer.