Ticagrelor

Ticagrelor with or without Aspirin in High-Risk Patients after PCI
R. Mehran, U. Baber, S.K. Sharma, D.J. Cohen, D.J. Angiolillo, C. Briguori,
J.Y. Cha, T. Collier, G. Dangas, D. Dudek, V. Džavík, J. Escaned, R. Gil, P. Gurbel,
C.W. Hamm, T. Henry, K. Huber, A. Kastrati, U. Kaul, R. Kornowski, M. Krucoff,
V. Kunadian, S.O. Marx, S.R. Mehta, D. Moliterno, E.M. Ohman, K. Oldroyd,
G. Sardella, S. Sartori, R. Shlofmitz, P.G. Steg, G. Weisz, B. Witzenbichler,
Y. Han, S. Pocock, and C.M. Gibson

BACKGROUND
Monotherapy with a P2Y12 inhibitor after a minimum period of dual antiplatelet therapy is an emerging approach to reduce the risk of bleeding after percutaneous coronary intervention (PCI).
METHODS
In a double-blind trial, we examined the effect of ticagrelor alone as compared with ticagrelor plus aspirin with regard to clinically relevant bleeding among patients who were at high risk for bleeding or an ischemic event and had undergone PCI. After 3 months of treatment with ticagrelor plus aspirin, patients who had not had a major bleeding event or ischemic event continued to take ticagrelor and were randomly assigned to receive aspirin or placebo for 1 year. The primary end point was Bleeding Academic Research Consortium (BARC) type 2, 3, or 5 bleeding. We also evaluated the composite end point of death from any cause, nonfatal myocar- dial infarction, or nonfatal stroke, using a noninferiority hypothesis with an ab- solute margin of 1.6 percentage points.
RESULTS
We enrolled 9006 patients, and 7119 underwent randomization after 3 months. Between randomization and 1 year, the incidence of the primary end point was 4.0% among patients randomly assigned to receive ticagrelor plus placebo and 7.1% among patients assigned to receive ticagrelor plus aspirin (hazard ratio, 0.56; 95% confidence interval [CI], 0.45 to 0.68; P<0.001). The difference in risk between the groups was similar for BARC type 3 or 5 bleeding (incidence, 1.0% among patients receiving ticagrelor plus placebo and 2.0% among patients receiving tica- grelor plus aspirin; hazard ratio, 0.49; 95% CI, 0.33 to 0.74). The incidence of death from any cause, nonfatal myocardial infarction, or nonfatal stroke was 3.9% in both groups (difference, −0.06 percentage points; 95% CI, −0.97 to 0.84; hazard ratio, 0.99; 95% CI, 0.78 to 1.25; P<0.001 for noninferiority). CONCLUSIONS Among high-risk patients who underwent PCI and completed 3 months of dual antiplatelet therapy, ticagrelor monotherapy was associated with a lower incidence of clinically relevant bleeding than ticagrelor plus aspirin, with no higher risk of death, myocardial infarction, or stroke. (Funded by AstraZeneca; TWILIGHT ClinicalTrials.gov number, NCT02270242.)  The authors’ full names, academic de- grees, and affiliations are listed in the Appendix. Address reprint requests to Dr. Mehran at the Icahn School of Medicine at Mount Sinai, 1 Gustav L. Levy Place, Box 1030, New York, NY 10029, or at [email protected]. Drs. Mehran and Baber contributed equally to this article. This article was published on September 26, 2019, at NEJM.org. DOI: 10.1056/NEJMoa1908419 Copyright © 2019 Massachusetts Medical Society. n engl j med nejm.org 1 T he new engl a nd jour na l of medicine A mong patients who have an acute coronary syndrome or who have under- gone percutaneous coronary intervention (PCI), the risk of thrombotic events is lower with dual antiplatelet therapy with aspirin and a P2Y12 receptor inhibitor than with aspirin alone.1 Even with dual antiplatelet therapy, the risk of adverse events remains unacceptably high among pa- tients with enhanced thrombotic risk due to clinical factors (e.g., diabetes mellitus) or angio- graphic factors (e.g., complex coronary artery disease).2-5 The use of more potent P2Y12 inhibi- tors or extension of the duration of dual anti- platelet therapy lowers residual ischemic risk among such patients but increases bleeding.6-9 Although previously considered relatively benign, post-PCI bleeding has been shown to be associ- ated with a substantial and durable risk of death, approximating or even exceeding that associated with myocardial infarction.2,10,11 Addressing the clinical imperatives of lower- ing the risk of bleeding while preserving ische- mic benefit requires therapeutic strategies that decouple thrombotic risk from hemorrhagic risk. One approach involves shortening the duration of dual antiplatelet therapy through early with- drawal of P2Y12 inhibition.12 Although several studies have shown the feasibility of this ap- proach, they generally have enrolled predomi- nantly low-risk patients and were underpowered for ischemic events.13-15 Reducing the duration of aspirin therapy may allow for more prolonged use of potent P2Y12 inhibitors while avoiding aspirin-related bleeding risk, particularly with respect to gastrointestinal toxicity.16 We designed the Ticagrelor with Aspirin or Alone in High-Risk Patients after Coronary Intervention (TWILIGHT) trial to test the hypothesis that in patients un- dergoing PCI who are at high risk for ischemic or hemorrhagic complications and who have completed a 3-month course of dual antiplatelet therapy with ticagrelor plus aspirin, continued treatment with ticagrelor monotherapy would be superior to ticagrelor plus aspirin with respect to clinically relevant bleeding and would not lead to ischemic harm. Trial Design and Oversight We conducted this randomized, placebo-controlled trial in 187 sites across 11 countries. The trial rationale and design have been described previ- 2 n engl j med  ously.17 The Icahn school of Medicine at Mount sinai designed and sponsored the trial, which was supported by an investigator-initiated grant from AstraZeneca. The executive and steering committees were responsible for trial conduct, the integrity of the data analysis, and the report- ing of results. National regulatory agencies and institutional review boards or ethics committees of participating centers approved the trial proto- col, which is available with the full text of this article at NEJM.org. An independent data and safety monitoring board provided external over- sight to ensure the safety of the trial partici- pants. All the authors vouch for the adherence of the trial to the protocol, and the first, second, and last authors vouch for the accuracy and com- pleteness of the data. The committee members and participating investigators are listed in Table s1 in the supplementary Appendix, available at NEJM.org. AstraZeneca provided financial sup- port and supplied ticagrelor for the trial but had no role in the design, collection, analysis, or interpretation of the data, in the preparation of the manuscript, or in the decision to submit the manuscript for publication. Trial population Patients who underwent successful PCI with at least one locally approved drug-eluting stent and whom the treating clinician intended to dis- charge with a regimen of ticagrelor plus aspirin were eligible to participate. Patients also had to have at least one additional clinical feature and one angiographic feature associated with a high risk of ischemic or bleeding events.2-5,17 The clinical criteria for high risk were an age of at least 65 years, female sex, troponin-positive acute coronary syndrome, established vascular disease, diabetes mellitus that was being treated with medication, and chronic kidney disease. Angio- graphic criteria included multivessel coronary artery disease, a total stent length of more than 30 mm, a thrombotic target lesion, a bifurcation lesion treated with two stents, an obstructive left main or proximal left anterior descending le- sion, and a calcified target lesion treated with atherectomy. Key exclusion criteria included pre- sentation with sT-segment elevation myocardial infarction, cardiogenic shock, ongoing long-term treatment with oral anticoagulants, or contrain- dication to aspirin or ticagrelor. (Tables s2 and s3 show all the criteria and their relation to bleeding and ischemic risks.) nejm.org Ticagrelor with or without aspirin after PCi Trial Regimen All enrolled patients received open-label ticagrel- or (90 mg twice daily) and enteric-coated aspirin (81 to 100 mg daily) after the index PCI. At 3 months after hospital discharge, patients who had not had a major bleeding event (see below) or an ischemic event (stroke, myocardial infarc- tion, or coronary revascularization) were eligible to be randomly assigned in a 1:1 ratio in a double-blind fashion to receive aspirin or match- ing placebo for an additional 12 months along with continuation of open-label ticagrelor treat- ment. For the determination of eligibility for randomization, we defined a major bleeding event as Bleeding Academic Research Consor- tium (BARC) type 3b or higher. BARC bleeding types range from 0 (no bleeding) to 5 (fatal bleeding); type 3b indicates overt bleeding lead- ing to a decrease in hemoglobin level of at least 5 mg per deciliter, cardiac tamponade, surgical intervention, or intravenous treatment with vaso- active drugs. Nonadherence to treatment with ticagrelor or aspirin rendered patients ineligible for randomization. A 3-month course of dual antiplatelet therapy before randomization was considered sufficient on the basis of trials that have suggested equipoise for such a duration.13,18 Randomization was performed with a secure Web-based system; an independent statistician who was not involved with the trial generated the randomization sequence, which was strati- fied according to site with randomly varying block sizes of 4, 6, and 8. Follow-up was per- formed by telephone at 1 month after random- ization and in person at 6 and 12 months after randomization. Adherence was assessed with manual pill counts, and nonadherence was clas- sified according to the underlying reason, as de- scribed previously.19 After 12 months of protocol- mandated therapy, patients were switched to a standard-of-care antiplatelet regimen at the dis- cretion of their treating physician, followed by final telephone follow-up 3 months later. End points The primary end point was the first occurrence of BARC type 2, 3, or 5 bleeding between ran- domization and 1 year in a time-to-event analy- sis. The key secondary end point was the first occurrence of death from any cause, nonfatal myocardial infarction, or nonfatal stroke in a time-to-event analysis. secondary bleeding end points included BARC type 3 or 5 bleeding20; n engl j med Thrombolysis in Myocardial Infarction (TIMI) major or minor bleeding21; Global Utilization of streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUsTO) moder- ate, severe, or life-threatening bleeding22; or major bleeding as defined by the International society on Thrombosis or Haemostasis (IsTH).23 Other secondary end points included death from cardiovascular causes, myocardial infarction, ischemic stroke, and definite or probable stent thrombosis. Myocardial infarction was defined according to the third universal definition,24 and revascularization and stent thrombosis were clas- sified according to the Academic Research Con- sortium.25 Table s4 lists the primary and all secondary end points and their associated defi- nitions. All clinical events were adjudicated by an external independent committee, the mem- bers of which were unaware of the treatment- group assignments. Statistical Analysis The sample-size and power calculation was based on a superiority assumption for the primary end point of BARC type 2, 3, or 5 bleeding. Assum- ing a bleeding incidence of 4.5% at 1 year with ticagrelor plus aspirin, we chose a sample size of 8200, which provided 80% power to detect a 28% lower incidence in the ticagrelor-plus-placebo group with a type I error rate of 0.05. The key secondary end point (composite of death from any cause, nonfatal myocardial infarction, or non- fatal stroke) was evaluated with the use of a prespecified noninferiority hypothesis. Under the assumption of an incidence of 8.0% at 1 year in the ticagrelor-plus-aspirin group, a sample size of 8200 provided 80% power to rule out an ab- solute difference in risk of 1.6 percentage points, with a one-sided type I error rate of 0.025. This margin is consistent with those in other trials that have evaluated pharmacologic and device- based interventions within a noninferiority framework.13,26 The cumulative incidence of the primary and secondary end points was estimated by the Kaplan–Meier method. Data from patients who had not had a primary end-point event between randomization and 1 year were censored at the time of death, the time of last known contact, or 365 days, whichever came first. Hazard ratios and 95% confidence intervals were generated with Cox proportional-hazards models. Absolute differences and 95% confidence intervals for nejm.org 3 T he new engl a nd jour na l of medicine primary and key secondary end points at 1 year were calculated with Kaplan–Meier estimates and Greenwood standard errors.27 Primary analy- ses of bleeding and ischemic end points were performed in the intention-to-treat and per-proto- col populations, respectively. Patients who under- went randomization and did not fulfill enroll- ment criteria, were not eligible for randomization, or never received protocol-mandated therapy were excluded from the per-protocol analysis. patient Characteristics From July 2015 through December 2017, a total of 9006 patients were enrolled after PCI, and 7119 were randomly assigned 3 months later to receive ticagrelor plus placebo or ticagrelor plus aspirin (intention-to-treat population). The last patient underwent randomization in April 2018, and the database was locked in July 2019. Details regarding the reasons for 1887 enrolled patients not undergoing randomization, the baseline clin- ical and procedural characteristics among patients who did and those who did not undergo ran- domization, and adverse events among patients who did not undergo randomization are provided in Tables s5 through s9. Ascertainment of the primary end point was complete in 98.4% of the patients who under- went randomization, and data on vital status were obtained in 99.7% (Fig. 1). Demographic, clinical, and procedural characteristics were well balanced between the treatment groups; the mean age was 65 years, 23.8% of the patients were female, 36.8% had diabetes mellitus, and 64.8% underwent PCI for an acute coronary syndrome indication (29.8% with non–sT-segment elevation myocardial infarction) (Table 1 and Table s10). Adherence to ticagrelor treatment 1 year after randomization was similar in the ticagrelor-plus- placebo group and the ticagrelor-plus-aspirin group (87.1% and 85.9%, respectively) (Fig. s1).19 Bleeding Table 2 shows the incidences of primary (BARC) and secondary (TIMI, GUsTO, and IsTH) bleed- ing end points, and Figure 2 shows the Kaplan– Meier curves for the primary end point. The primary end point occurred in 141 patients (4.0%) who received ticagrelor plus placebo, as compared with 250 patients (7.1%) who received  ticagrelor plus aspirin (hazard ratio, 0.56; 95% confidence interval [CI], 0.45 to 0.68; P<0.001), for a difference in risk of −3.08 percentage points (95% CI, −4.15 to −2.01). The incidence of BARC type 3 or 5 bleeding was 1.0% in the group that received ticagrelor plus placebo and 2.0% in the group that received ticagrelor plus aspirin (hazard ratio, 0.49; 95% CI, 0.33 to 0.74). The treatment effect for the primary end point was consistent across predefined subgroups (Fig. s2). Ischemic Events Ischemic events were analyzed in the per-proto- col population, which included the 7039 patients who underwent randomization and had no major deviations from the protocol (3524 who received ticagrelor plus placebo and 3515 who received ticagrelor plus aspirin). The key secondary com- posite end point of death from any cause, non- fatal myocardial infarction, or nonfatal stroke occurred in 135 patients (3.9%) who received tica- grelor plus placebo and in 137 patients (3.9%) who received ticagrelor plus aspirin (hazard ratio, 0.99; 95% CI, 0.78 to 1.25), for a difference in risk of −0.06 percentage points (95% CI, −0.97 to 0.84) (Fig. 3). The incidence of death from any cause was similar in group that received ticagrel- or plus placebo and the group that received tica- grelor plus aspirin (1.0% and 1.3%, respectively), as were the incidences of myocardial infarction (2.7% in both groups) and definite or probable stent thrombosis (0.4% and 0.6%) (Table 2). There were 16 instances of ischemic stroke in the group that received ticagrelor plus placebo and 8 instances of ischemic stroke in the group that received ticagrelor plus aspirin (0.5% and 0.2% of patients, respectively). The effect of tica- grelor monotherapy on the key secondary end point was consistent across predefined subgroups (Fig. s3). Additional Analyses Landmark analyses of the period from 15 to 18 months after PCI showed similar incidences of BARC type 2, 3, or 5 bleeding in the group that received ticagrelor plus placebo and the group that received ticagrelor plus aspirin (0.7% and 0.5%, respectively; hazard ratio, 1.24; 95% CI, 0.64 to 2.40); the incidence of death, myocardial infarction, or stroke was also similar in the two groups (0.9% and 1.1%; hazard ratio, 0.84; 95% 4 n engl j med  nejm.org Ticagrelor with or without aspirin after PCi CI, 0.51 to 1.40). sensitivity analyses conducted with an imputation-based approach to account for the 112 patients who were lost to clinical follow-up yielded similar effects for the primary and key secondary end points.28 (Details of these analyses are provided in Tables s11 and s12 and Fig. s4.) Our trial was designed to examine the effect of withdrawing treatment with aspirin while con- tinuing treatment with ticagrelor alone after 3 months of dual antiplatelet therapy in patients who received drug-eluting stents and were at high risk for bleeding or ischemic events. Ticagrelor monotherapy was associated with a 44% lower risk of BARC type 2, 3, or 5 bleeding over 1 year than ticagrelor plus aspirin (absolute difference in risk, 3.1 percentage points). The bleeding- related benefits of ticagrelor monotherapy ex- tended to more severe BARC type 3 or 5 bleeds and persisted when alternative bleeding scales were considered. In this trial, there was no evi- dence of a higher risk of death, myocardial in- farction, or stroke among patients who received ticagrelor monotherapy than among those who received ticagrelor plus aspirin. The treatment effect with respect to both bleeding and ische- mic end points was consistent across subgroups. In aggregate, these results show that a transition to an antiplatelet strategy of treatment with tica- grelor alone after a 3-month course of dual anti- platelet therapy in high-risk patients who had undergone PCI provided a clinical benefit of less bleeding without ischemic harm. n engl j med nejm.org 5 T he new engl a nd jour na l of medicine Table 1. Baseline Characteristics of the Patients Who Underwent Randomization.* Characteristic Ticagrelor plus Placebo (N = 3555) Ticagrelor plus Aspirin (N = 3564) Age — yr 65.2±10.3 65.1±10.4 Female sex — no. (%) 846 (23.8) 852 (23.9) Nonwhite race — no. (%)† 1110 (31.2) 1086 (30.5) Body-mass index‡ 28.6±5.5 28.5±5.6 Enrolling region — no. (%) North America 1484 (41.7) 1488 (41.8) Europe 1251 (35.2) 1258 (35.3) Asia 820 (23.1) 818 (23.0) Diabetes mellitus — no. (%) 1319 (37.1) 1301 (36.5) Diabetes treated with insulin — no. (%) 335 (9.4) 374 (10.5) Chronic kidney disease — no./total no. (%)§ 572/3410 (16.8) 573/3425 (16.7) Anemia — no./total no. (%) 675/3405 (19.8) 654/3423 (19.1) Current smoker — no./total no. (%) 726/3553 (20.4) 822/3562 (23.1) Hypercholesterolemia — no. (%) 2157 (60.7) 2146 (60.2) Hypertension — no./total no. (%) 2580/3555 (72.6) 2574/3563 (72.2) Peripheral arterial disease — no. (%) 245 (6.9) 244 (6.8) Previous myocardial infarction — no. (%) 1020 (28.7) 1020 (28.6) Previous PCI — no. (%) 1502 (42.3) 1496 (42.0) Previous CABG — no./total no. (%) 362/3554 (10.2) 348/3564 (9.8) Multivessel coronary artery disease — no. (%) 2272 (63.9) 2194 (61.6) Previous major bleeding event — no. (%) 31 (0.9) 32 (0.9) Indication for PCI — no./total no. (%) Asymptomatic 234/3554 (6.6) 223/3563 (6.3) Stable angina 1047/3554 (29.5) 999/3563 (28.0) Unstable angina 1249/3554 (35.1) 1245/3563 (34.9) NSTEMI 1024/3554 (28.8) 1096/3563 (30.8) * Plus–minus values are means ±SD. Percentages may not total 100 because of rounding. CABG denotes coronary artery bypass graft, NSTEMI non–ST-segment elevation myocardial infarction, and PCI percutaneous coronary intervention. † Race was reported by the patient. ‡ The body-mass index is the weight in kilograms divided by the square of the height in meters. § Chronic kidney disease was defined as an estimated glomerular filtration rate of less than 60 ml per minute per 1.73 m2 of body-surface area. Two previous studies showed that among pa- tients who had undergone PCI and were at rela- tively low risk for ischemic events, clopidogrel monotherapy after 1 to 3 months of dual anti- platelet therapy was associated with a signifi- cantly lower incidence of bleeding than clopido- grel plus aspirin, without an apparent difference in ischemic risk.29,30 The modest size of those studies, as well as the low-risk nature of the patient population, precluded conclusive infer- ence regarding the effect of clopidogrel mono- therapy on ischemic end points. Distinct from these trials, we enrolled a larger population of patients who more commonly had both clinical and angiographic high-risk criteria and were treated with a more potent P2Y12 inhibitor, tica- grelor. In contrast to our findings, the findings of the GLOBAL LEADERs trial showed that 1 month of dual antiplatelet therapy followed by ticagrelor monotherapy for an additional 23 months was not associated with a lower incidence of bleeding 6 n engl j med  nejm.org Ticagrelor with or without aspirin after PCi Table 2. Bleeding and Ischemic Events 1 Year after Randomization.* Variable Ticagrelor plus Ticagrelor plus Placebo Aspirin (N = 3555) (N = 3564) no. of patients (%)‡ Hazard Ratio (95% CI)† P Value Bleeding end points Primary end point: BARC type 2, 3, or 5§ 141 (4.0) 250 (7.1) 0.56 (0.45–0.68) <0.001¶ BARC type 3 or 5§ 34 (1.0) 69 (2.0) 0.49 (0.33–0.74) TIMI minor or major 141 (4.0) 250 (7.1) 0.56 (0.45–0.68) GUSTO moderate or severe 26 (0.7) 49 (1.4) 0.53 (0.33–0.85) ISTH major 39 (1.1) 72 (2.1) 0.54 (0.37–0.80) Ischemic end points Death from any cause, nonfatal myocardial infarction, or nonfatal stroke 135 (3.9) 137 (3.9) 0.99 (0.78–1.25) <0.001‖ Death from cardiovascular causes, nonfatal myocardial infarction, or nonfatal ischemic stroke 126 (3.6) 130 (3.7) 0.97 (0.76–1.24) Death from any cause 34 (1.0) 45 (1.3) 0.75 (0.48–1.18) Death from cardiovascular causes 26 (0.8) 37 (1.1) 0.70 (0.43–1.16) Myocardial infarction 95 (2.7) 95 (2.7) 1.00 (0.75–1.33) Ischemic stroke 16 (0.5) 8 (0.2) 2.00 (0.86–4.67) Stent thrombosis, definite or probable 14 (0.4) 19 (0.6) 0.74 (0.37–1.47) * Bleeding end points were evaluated in the intention-to-treat population (the 7119 patients who underwent randomization), and ischemic end points were evaluated in the per-protocol population (the 7039 patients who underwent randomiza- tion and had no major deviations from the protocol [3524 who received ticagrelor plus placebo and 3515 who received ticagrelor plus aspirin]). All primary and secondary end points and their associated definitions are listed in Table S4. GUSTO denotes Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries, ISTH International Society on Thrombosis and Hemostasis, and TIMI Thrombolysis in Myocardial Infarction. † The 95% confidence intervals for secondary end points have not been adjusted for multiplicity, and therefore inferences drawn from these intervals may not be reproducible. ‡ Event percentages are Kaplan–Meier estimates of the incidence of the end point at 12 months after randomization. § Bleeding Academic Research Consortium (BARC) types range from 0 (no bleeding) to 5 (fatal bleeding). ¶The difference in the risk of the primary end point of BARC 2, 3, or 5 bleeding was −3.08 percentage points (95% CI, −4.15 to −2.01). ‖ The difference in the risk of the key secondary end point of death from any cause, nonfatal myocardial infarction, or nonfatal stroke was −0.06 percentage points (95% CI, −0.97 to 0.84). For the key secondary end point, the upper limit of the 95% confidence interval for the difference indicated noninferiority (P<0.001). than a conventional antiplatelet strategy.31 These results may be attributable to differences in trial design (double-blind vs. open-label), patient case mix (high-risk patients vs. all comers), duration of therapy after randomization (12 months vs. 23 months), comparator regimens (ticagrelor plus aspirin vs. dual antiplatelet therapy followed by aspirin), bleeding ascertainment (adjudicated vs. site-reported), or protocol adherence. Conse- quently, any putative bleeding-related advantage associated with the withdrawal of aspirin therapy may have been attenuated in GLOBAL LEADERs. For instance, ticagrelor monotherapy was associ- ated with a nonsignificant 14% lower incidence of BARC type 3 or 5 bleeding at 1 year in GLOBAL LEADERs, whereas a 51% lower incidence was observed in our trial. To be included in our trial, patients had to have clinical and angiographic factors associated with a high risk of either bleeding or ischemic events after PCI, design features that reflect the primary and key secondary end points of the trial. With respect to bleeding, patients at high risk are most likely to have a benefit from re- duced exposure to antiplatelet therapy. With regard to ischemic events, we enrolled patients with a high risk of such events to identify signals of harm after withdrawal of aspirin therapy. Although n engl j med nejm.org 7 T he new engl a nd jour na l of medicine 100 Cumulative Incidence (%) 80 60 40 20  10 Hazard ratio, 0.56 (95% CI, 0.45–0.68) P<0.001 8 Ticagrelor plus aspirin 6 4 2 Ticagrelor plus placebo 0 0 3 6 9 12  pirin and ticagrelor were eligible for enrollment, which resulted in a trial population with a level of ischemic and bleeding risk that reflects the overall trial design and clinical preferences for ticagrelor use. Major adverse events occurring early after PCI and nonadherence to dual antiplatelet therapy precluded randomization at 90 days, criteria that led to the population of patients who underwent randomization having a clinical and angiographic No. at Risk 0 0 3 6 9 12 Months since Randomization profile distinct from that of the initially enrolled participants. Nonetheless, several high-risk char- acteristics (e.g., diabetes mellitus and long stent length) remained prevalent among the patients Ticagrelor plus aspirin Ticagrelor plus placebo 3564 3555 3454 3474 3357 3424 3277 3366 3213 3321 who underwent randomization. Moreover, the in- cidences of bleeding and ischemic events at 1 year Figure 2. Kaplan–Meier Estimates of the Incidence of BARC Type 2, 3, or 5 Bleeding 1 Year after Randomization (Intention-to-Treat Population). The hazard ratio shown is for ticagrelor plus placebo versus ticagrelor plus aspirin. Bleeding Academic Research Consortium (BARC) types range from 0 (no bleeding) to 5 (fatal bleeding). The inset shows the same data on an expanded y axis. CI denotes confidence interval. in these patients were similar to or higher than those reported in trials in which all events from PCI onward were considered, thereby substantiat- ing the high-risk nature of our trial population.29-31 Although guidelines recommend ticagrelor in the context of acute coronary syndrome alone, 33% of the trial participants were in stable con- Cumulative Incidence (%) dition at the time of enrollment.1 Potential rea- sons for the inclusion of such patients by trial investigators may have included a perceived lack of benefit with clopidogrel or clinical equipoise with regard to P2Y12 inhibitor choice in high-risk patients who are stable after PCI. Corroborating such tendencies, findings from usual-care regis- tries show that in current practice, more than 10% of patients who have undergone PCI and are treated with ticagrelor initially present with a non–acute coronary syndrome indication.32,33 From a clinical standpoint, our results suggest that ticagrelor monotherapy may be a suitable anti- platelet strategy to lower the risk of bleeding while simultaneously preserving ischemic bene- fit in patients who have undergone PCI and are characterized by a minimum threshold of risk. These effects appear consistent in patients whose condition is either stable or acute. Whether the findings would be similar in a lower-risk popu- lation or if a different antithrombotic regimen were used remains unknown. The limitations of our trial include the lack of power to detect differences in the risk of impor- most of the prespecified criteria are associated with excess thrombosis (e.g., diabetes mellitus), others are linked to both types of events (e.g., renal impairment). Moreover, only patients whom a clinician had already decided to treat with as- tant yet rare clinical events, such as stent throm- bosis and stroke. Although ischemic strokes were more common among patients who re- ceived placebo than among those who continued to receive aspirin, only 24 such events occurred 8 n engl j med  nejm.org Ticagrelor with or without aspirin after PCi during the trial, thereby precluding conclusive inference for this end point. Other studies have shown that P2Y12 inhibitor monotherapy is not associated with a higher risk of cerebrovascular events than dual antiplatelet therapy.29,30,34 Re- sults from this trial may not be generalizable to all patients who have undergone PCI, given the requirement in our trial for both high-risk (clini- cal and angiographic) features and a willingness to be treated with ticagrelor. In addition, the observed treatment effects do not apply to all enrolled participants but rather to those patients who were able to take 3 months of dual anti- platelet therapy without any major adverse events. Our primary end point included bleeding events of varying severity, which may have altered the risk–benefit calculation for considering ticagrel- or monotherapy. A lower-than-expected incidence of the composite end point of death, myocardial infarction, or stroke may have biased our results for this key secondary end point toward the null. Our trial showed that in high-risk patients who had undergone PCI and were treated with ticagrelor and aspirin for 3 months, an anti- platelet strategy of continuing ticagrelor alone resulted in substantially less bleeding than tica- grelor plus aspirin, without leading to ischemic harm over a period of 1 year. A data sharing statement provided by the authors is available with the full text of this article at NEJM.org. supported by AstraZeneca. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. We thank Rishi Chandiramani, M.D., of the Icahn school of Medicine at Mount sinai, and Alexandra Howson, Ph.D., of Thistle Editorial, for assistance with preparation of an earlier version of the manuscript (funded by the Icahn school of Medi- cine at Mount sinai). Appendix The authors’ full names and academic degrees are as follows: Roxana Mehran, M.D., Usman Baber, M.D., samin K. sharma, M.D., David J. Cohen, M.D., Dominick J. Angiolillo, M.D., Ph.D., Carlo Briguori, M.D., Ph.D., Jin Y. Cha, B.s., Timothy Collier, M.sc., George Dangas, M.D., Ph.D., Dariusz Dudek, M.D., Ph.D., Vladimír Džavík, M.D., Javier Escaned, M.D., Ph.D., Robert Gil, M.D., Ph.D., Paul Gurbel, M.D., Christian W. Hamm, M.D., Timothy Henry, M.D., Kurt Huber, M.D., Adnan Kastrati, M.D., Upendra Kaul, M.D., Ran Kornowski, M.D., Mitchell Krucoff, M.D., Vijay Kunadian, M.B., B.s., M.D., steven O. Marx, M.D., shamir R. Mehta, M.D., David Moliterno, M.D., E. Magnus Ohman, M.D., Keith Oldroyd, M.B., Ch.B., M.D., Gennaro sardella, M.D., samantha sartori, Ph.D., Rich- ard shlofmitz, M.D., P. Gabriel steg, M.D., Giora Weisz, M.D., Bernhard Witzenbichler, M.D., Ya-ling Han, M.D., Ph.D., stuart Pocock, Ph.D., and C. Michael Gibson, M.D. The authors’ affiliations are as follows: the Zena and Michael A. Wiener Cardiovascular Institute, Mount sinai Hospital (R.M., U.B., J.Y.C., G.D., s.s.), Mount sinai Health system (s.K.s.), Columbia University Medical Center (s.O.M.), Icahn school of Medicine at Mount sinai (s.s.), and Montefiore Medical Center (G.W.), New York, and st. Francis Hospital, Roslyn (R.s.) — all in New York; Kan- sas City, Missouri (D.J.C.); the University of Florida–shands, Jacksonville (D.J.A.); Clinica Mediterranea, Naples (C.B.), and Policlinico Umberto I University, Rome (G.s.) — both in Italy; the London school of Hygiene and Tropical Medicine, London (T.C., s.P.), Institute of Cellular Medicine, Faculty of Medical sciences, Newcastle University, and Cardiothoracic Centre, Freeman Hospital, Newcastle upon Tyne Hospitals NHs Foundation Trust, Newcastle upon Tyne (V.K.), and the West of scotland Heart and Lung Centre, Golden Jubilee National Hospital, Glasgow (K.O.) — all in the United Kingdom; the 2nd Department of Cardiology Jagiellonian University Medical College, Krakow (D.D.), and the Department of Invasive Cardiology, Center of Postgraduate Medical Education, Central Clinical Hospi- tal of the Ministry of Interior and Administration, Warsaw (R.G.) — both in Poland; Research and Innovation in Interventional Cardiol- ogy and Cardiac Intensive Care, Peter Munk Cardiac Centre, University Health Network, Toronto (V.D.), and Hamilton Health sciences, Hamilton, ON (s.R.M.) — both in Canada; Instituto de Investigación sanitaria del Hospital Clínico san Carlos and Complutense Uni- versity, Madrid (J.E.); sinai Hospital of Baltimore system, Baltimore (P.G.); Kerckhoff Clinic, Bad Nauheim (C.W.H.), Deutsches Herzzen- trum München, Munich (A.K.), and Helios Amper-Klinikum, Dachau (B.W.) — all in Germany; the Carl and Edyth Lindner Center for Research and Education at the Christ Hospital, Cincinnati (T.H.); Wilhelminenhospital, Vienna (K.H.); Batra Hospital and Medical Research Centre, New Delhi, India (U.K.); Rabin Medical Center, Petach Tikva, Israel (R.K.); Duke University Medical Center–Duke Clinical Research Institute, Durham, NC (M.K., E.M.O.); the University of Kentucky, Lexington (D.M.); Groupe Hospitalier Bichat–Claude- Bernard, Paris (P.G.s.); shenyang North Hospital, shenyang, China (Y.H.); and Beth Israel Deaconess Medical Center, Boston (C.M.G.). References ⦁ Capodanno D, Alfonso F, Levine GN, Valgimigli M, Angiolillo DJ. ACC/AHA ver- sus EsC guidelines on dual antiplatelet therapy: JACC guideline comparison. J Am Coll Cardiol 2018;72:2915-31. ⦁ Généreux P, Giustino G, Witzenbichler B, et al. Incidence, predictors, and impact of post-discharge bleeding after percuta- neous coronary intervention. J Am Coll Cardiol 2015;66:1036-45. ⦁ Costa F, van Klaveren D, James s, et al. Derivation and validation of the predicting bleeding complications in patients under-  going stent implantation and subsequent dual antiplatelet therapy (PRECIsE-DAPT) score: a pooled analysis of individual- patient datasets from clinical trials. Lan- cet 2017;389:1025-34. ⦁ Yeh RW, secemsky EA, Kereiakes DJ, et al. Development and validation of a pre- diction rule for benefit and harm of dual antiplatelet therapy beyond 1 year after percutaneous coronary intervention. JAMA 2016;315:1735-49. ⦁ Baber U, Mehran R, Giustino G, et al. Coronary thrombosis and major bleeding  after PCI with drug-eluting stents: risk scores from PARIs. J Am Coll Cardiol 2016;67:2224-34. ⦁ Bonaca MP, Bhatt DL, Cohen M, et al. Long-term use of ticagrelor in patients with prior myocardial infarction. N Engl J Med 2015;372:1791-800. ⦁ Mauri L, Kereiakes DJ, Yeh RW, et al. Twelve or 30 months of dual antiplatelet therapy after drug-eluting stents. N Engl J Med 2014;371:2155-66. ⦁ Wallentin L, Becker RC, Budaj A, et al. Ticagrelor versus clopidogrel in patients n engl j med nejm.org 9 T he new engl a nd jour na l of medicine with acute coronary syndromes. N Engl J Med 2009;361:1045-57. ⦁ Wiviott sD, Braunwald E, McCabe CH, et al. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2007;357:2001-15. ⦁ Baber U, Dangas G, Chandrasekhar J, et al. Time-dependent associations between actionable bleeding, coronary thrombotic events, and mortality following percuta- neous coronary intervention: results from the PARIs Registry. JACC Cardiovasc In- terv 2016;9:1349-57. ⦁ Valgimigli M, Costa F, Lokhnygina Y, et al. Trade-off of myocardial infarction vs. bleeding types on mortality after acute coronary syndrome: lessons from the Thrombin Receptor Antagonist for Clini- cal Event Reduction in Acute Coronary syndrome (TRACER) randomized trial. Eur Heart J 2017;38:804-10. ⦁ Moon JY, Franchi F, Rollini F, Angio- lillo DJ. Evolution of coronary stent tech- nology and implications for duration of dual antiplatelet therapy. Prog Cardiovasc Dis 2018;60:478-90. ⦁ Feres F, Costa RA, Abizaid A, et al. Three vs twelve months of dual antiplatelet therapy after zotarolimus-eluting stents: the OPTIMIZE randomized trial. JAMA 2013;310:2510-22. ⦁ Kim BK, Hong MK, shin DH, et al. A new strategy for discontinuation of dual antiplatelet therapy: the REsET Trial (REal safety and Efficacy of 3-month dual antiplatelet Therapy following Endeavor zotarolimus-eluting stent implantation). J Am Coll Cardiol 2012;60:1340-8. ⦁ Valgimigli M, Campo G, Monti M, et al. short- versus long-term duration of dual- antiplatelet therapy after coronary stent- ing: a randomized multicenter trial. Cir- culation 2012;125:2015-26. ⦁ Capodanno D, Mehran R, Valgimigli M, et al. Aspirin-free strategies in cardio- vascular disease and cardioembolic stroke prevention. Nat Rev Cardiol 2018;15:480- 96. ⦁ Baber U, Dangas G, Cohen DJ, et al. Ticagrelor with aspirin or alone in high- risk patients after coronary intervention:  rationale and design of the TWILIGHT study. Am Heart J 2016;182:125-34. ⦁ Palmerini T, Benedetto U, Bacchi- Reggiani L, et al. Mortality in patients treated with extended duration dual anti- platelet therapy after drug-eluting stent implantation: a pairwise and Bayesian net- work meta-analysis of randomised trials. Lancet 2015;385:2371-82. ⦁ Mehran R, Baber U, steg PG, et al. Cessation of dual antiplatelet treatment and cardiac events after percutaneous coronary intervention (PARIs): 2 year re- sults from a prospective observational study. Lancet 2013;382:1714-22. ⦁ Mehran R, Rao sV, Bhatt DL, et al. standardized bleeding definitions for car- diovascular clinical trials: a consensus report from the Bleeding Academic Re- search Consortium. Circulation 2011;123: 2736-47. ⦁ Bovill EG, Terrin ML, stump DC, et al. Hemorrhagic events during therapy with recombinant tissue-type plasminogen ac- tivator, heparin, and aspirin for acute myo- cardial infarction: results of the Throm- bolysis in Myocardial Infarction (TIMI), phase II trial. Ann Intern Med 1991;115: 256-65. ⦁ GUsTO investigators. An international randomized trial comparing four throm- bolytic strategies for acute myocardial infarction. N Engl J Med 1993;329:673-82. ⦁ Kaatz s, Ahmad D, spyropoulos AC, schulman s. Definition of clinically rele- vant non-major bleeding in studies of anti- coagulants in atrial fibrillation and venous thromboembolic disease in non- surgical patients: communication from the ssC of the IsTH. J Thromb Haemost 2015;13:2119-26. ⦁ Thygesen K, Alpert Js, Jaffe As, et al. Third universal definition of myocardial infarction. J Am Coll Cardiol 2012;60: 1581-98. ⦁ Cutlip DE, Windecker s, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation 2007;115:2344-51. ⦁ stone GW, Rizvi A, Newman W, et al. Everolimus-eluting versus paclitaxel-elut-  ing stents in coronary artery disease. N Engl J Med 2010;362:1663-74. ⦁ Com-Nougue C, Rodary C, Patte C. How to establish equivalence when data are censored: a randomized trial of treat- ments for B non-Hodgkin lymphoma. stat Med 1993;12:1353-64. ⦁ Reardon MJ, Van Mieghem NM, Popma JJ. surgical or transcatheter aortic-valve re- placement. N Engl J Med 2017;377:197-8. ⦁ Hahn JY, song YB, Oh JH, et al. Effect of P2Y12 inhibitor monotherapy vs dual anti- platelet therapy on cardiovascular events in patients undergoing percutaneous coro- nary intervention: the sMART-CHOICE randomized clinical trial. JAMA 2019;321: 2428-37. ⦁ Watanabe H, Domei T, Morimoto T, et al. Effect of 1-month dual antiplatelet therapy followed by clopidogrel vs 12- month dual antiplatelet therapy on cardio- vascular and bleeding events in patients receiving pci: the sTOPDAPT-2 random- ized clinical trial. JAMA 2019;321:2414-27. ⦁ Vranckx P, Valgimigli M, Jüni P, et al. Ticagrelor plus aspirin for 1 month, fol- lowed by ticagrelor monotherapy for 23 months vs aspirin plus clopidogrel or tica- grelor for 12 months, followed by aspirin monotherapy for 12 months after implan- tation of a drug-eluting stent: a multi- centre, open-label, randomised superiority trial. Lancet 2018;392:940-9. ⦁ Fan W, Plent s, Prats J, Deliargyris EN. Trends in P2Y12 inhibitor use in pa- tients referred for invasive evaluation of coronary artery disease in contemporary Us practice. Am J Cardiol 2016;117:1439-43. ⦁ Karve AM, seth M, sharma M, et al. Contemporary use of ticagrelor in inter- ventional practice (from Blue Cross Blue shield of Michigan Cardiovascular Con- sortium). Am J Cardiol 2015;115:1502-6.
⦁ Diener HC, Bogousslavsky J, Brass LM, et al. Aspirin and clopidogrel com- pared with clopidogrel alone after recent ischaemic stroke or transient ischaemic attack in high-risk patients (MATCH): randomised, double-blind, placebo-con- trolled trial. Lancet 2004;364:331-7.
Copyright © 2019 Massachusetts Medical Society.

10 n engl j med

nejm.org