Rationale, design and baseline characteristics of the effect of ticagrelor on health outcomes in diabetes mellitus patients Intervention study

In the setting of prior myocardial infarction, the oral antiplatelet ticagrelor added to aspirin reduced the risk of recurrent ischemic events, especially, in those with diabetes mellitus. Patients with stable coronary disease and diabetes are also at elevated risk and might benefit from dual antiplatelet therapy. The Effect of Ticagrelor on Health Outcomes in diabEtes Mellitus patients Intervention Study (THEMIS, NCT01991795) is a Phase 3b randomized, double‐blinded, placebo‐controlled trial of ticagrelor vs placebo, on top of low dose aspirin. Patients ≥50 years with type 2 diabetes receiving anti‐diabetic medications for at least 6 months with stable coronary artery disease as determined by a history of previous percutaneous coronary intervention, bypass grafting, or angiographic stenosis of ≥50% of at least one coronary artery were enrolled. Patients with known prior myocardial infarction (MI) or stroke were excluded. The primary efficacy endpoint is a composite of cardiovascular death, myocardial infarction, or stroke. The primary safety endpoint is Thrombolysis in Myocardial Infarction major bleeding. A total of 19 220 patients worldwide have been randomized and at least 1385 adjudicated primary efficacy endpoint events are expected to be available for analysis, with an expected average follow‐up of 40 months (maximum 58 months). Most of the exposure is on a 60 mg twice daily dose, as the dose was lowered from 90 mg twice daily partway into the study. The results may revise the boundaries of efficacy for dual antiplatelet therapy and whether it has a role outside acute coronary syndromes, prior myocardial infarction, or percutaneous coronary intervention.


| INTRODUCTION
Type 2 diabetes mellitus (DM) is a highly prevalent risk factor for coronary artery disease, with an incidence that is increasing worldwide. In those who have established atherosclerosis, the presence of DM further increases future risk of ischemic events in a synergistic fashion. 1 DM multiplies cardiovascular (CV) risk not only in those with prior ischemic events, but also in those with stable coronary artery disease.
Heightened platelet activity appears to be present in atherothrombotic patients with DM. Long-term dual antiplatelet therapy (DAPT) reduces CV event rates in patients with acute coronary syndromes, both in the short-and long term. [2][3][4][5][6][7][8][9][10][11] Dual antiplatelet therapy appears to have a particular benefit in patients with prior myocardial infarction (MI) and diabetes. [12][13][14][15] Whether that benefit extends to patients with diabetes and stable coronary artery disease without a history of prior MI remains a major unanswered question.
The Effect of Ticagrelor on Health Outcomes in diabEtes Mellitus patients Intervention Study (THEMIS) trial was designed to evaluate the potential benefits and risks of dual antiplatelet therapy with ticagrelor plus low dose aspirin vs placebo plus aspirin in patients with established stable coronary artery disease and DM treated with medications.
twice daily. A protocol amendment enabling a dose reduction was finalized in May 2015, resulting in that approximately 25% of the randomized patients started on ticagrelor 60 mg twice daily or matching placebo. Because the lower dose was introduced rather early in the study, it is expected that at end of the study, at least 75% of the exposure time will be on 60 mg twice daily. The primary efficacy endpoint is a composite of CV death, myocardial infarction, or stroke. The primary safety endpoint is Thrombolysis in Myocardial Infarction major bleeding. Table 2 provides further details about the pre-specified hierarchical testing of secondary endpoints. Appendix S1, Supporting Information provides the definitions. All endpoints were adjudicated by a blinded academic clinical endpoint committee. Partway through the trial, given an increasing appreciation of the importance of peripheral artery ischemic endpoints, these were also adjudicated by a blinded academic clinical endpoint committee. [15][16][17] To ensure data integrity, there was a firewall maintained between the clinical endpoint committee (CEC) and the data monitoring committee (DMC). Specifically, the DMC statistical data analysis center (SDAC), handling all unblinded study data, was located at Duke Clinical Research Institute (DCRI).
The primary efficacy endpoint will be tested at a 4.96% significance level (two-sided), adjusted for 1 planned efficacy interim analysis with family-wise error controlled at 5%. The one planned efficacy interim analysis occurred on the 29th of March 2017. The recommendation from the DMC was to continue the study according to the protocol. The estimated annual event rate in the placebo arm was 2.5%.
An effect size of 16% relative risk reduction was hypothesized, with 1385 primary endpoint events needed to provide 90% power. This resulted in an estimated sample size of 19 000 patients, randomized in a 1:1 ratio, with an average follow-up time of 40 months (maximum 58 months). Even with the ticagrelor dosage switch, a power of 90% in the study is maintained, as the main analysis is based on randomization to ticagrelor irrespective of dose. • Need for chronic oral anticoagulant therapy or chronic low-molecular-weight heparin (at venous thrombosis treatment not prophylaxis doses) • Known bleeding diathesis or coagulation disorder, or uncontrolled hypertension (defined as a systolic BP ≥180 mm Hg and/or diastolic BP ≥100 mmHg) • History of previous intracerebral bleed at any time, gastrointestinal bleed within 6 months prior to randomization, or major surgery within 30 days prior to randomization • Increased risk of bradycardic events (eg, known sick sinus syndrome, second or third degree AV block, or previous documented syncope suspected to be due to bradycardia) unless treated with a pacemaker • Known severe liver disease • Renal failure requiring dialysis • Pregnancy or lactation, and women of child-bearing potential not using reliable contraception • Concern for inability to comply with study procedures and/or follow-up, or any conditions (judged by the investigator) that may render the patient unable to complete the study • Any condition judged by the investigator that make participation unsafe or unsuitable, or any condition outside the atherothrombotic study area with a life expectancy <2 years • Participation in another clinical study with an investigational product within 28 days prior to enrolment, or previous randomization to an investigational product in another ongoing clinical study. Participation in any previous study with ticagrelor. Previous randomization in the present study  There are no sub-studies planned, but predefined subgroup analyses will explore efficacy and safety according to baseline characteristics, such as revascularization history, single vs multivessel coronary artery disease, duration of diabetes, glycemic control, anti-diabetic medications, age groups, and renal function.

| RESULTS
A total of 20 108 patients were enrolled. Of these, 19 271 were randomized, and because of closure of a single site secondary for inadequate adherence to good clinical practice in a different trial, 19 220 patients are expected to be available for analysis ( Figure 2).
The baseline characteristics and medications are described in

| DISCUSSION
The THEMIS trial was designed to address an important question in the management of patients with DM and stable coronary artery disease but no prior MI-does intensification of the antiplatelet regimen beyond aspirin reduce the risk of CV events? The trial is fully enrolled and well-powered to detect moderate relative risk reductions in the overall population studied. The sample size will hopefully allow examination of clinically logical and well-defined subgroups even if the overall trial does not meet its primary endpoint or shows marginal net clinical benefit. The evidence from THEMIS will complement that of several other trials which assessed the value of DAPT in patients with acute coronary syndromes or prior myocardial infarction. 2,7,11,12,15 When added to aspirin, ticagrelor specifically provides superior benefit compared with clopidogrel in patients with acute coronary syndromes 7 and compared with placebo in high-risk patients with prior myocardial infarction. 15 There is lack of clear benefit of ticagrelor monotherapy, compared with clopidogrel in peripheral artery disease 18 and compared with aspirin in ischemic stroke. 19 In the context of patients undergoing coronary stenting, there was no clear benefit to a strategy of DAPT for a month followed by ticagrelor monotherapy over a more conventional strategy of DAPT for a year followed by aspirin monotherapy, although additional study is ongoing in this regard. [20][21][22] There is evidence that platelet aggregation is enhanced in patients with DM compared with those without DM and that aspirin may have reduced efficacy in DM. 23 This appears to be because of several factors, including accelerated platelet turnover. Indeed, there is some evidence that twice daily administration of aspirin is more effective at inhibition of generation of thromboxane than once a day administration. 24,25 In addition, enteric coated aspirin is widely used and appears to be somewhat less effective than non-enteric coated aspirin in pharmacodynamic and pharmacokinetic analyses. 26 Recent data suggest that aspirin has only a very modest effect for primary prevention CV events in patients with DM. [27][28][29] Similarly, there is also evidence of reduced pharmacodynamic and pharmacokinetic efficacy of clopidogrel among patients with DM, in part because of less efficient metabolism of clopidogrel as a     Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; BMI, body mass index; CABG, coronary artery bypass grafting; CAD, coronary artery disease; eGFR, estimated glomerular filtration rate; HbA1c, glycated hemoglobin; IQR, interquartile range; MDRD, modification of diet in renal disease; PCI, percutaneous coronary intervention; PAD, peripheral artery disease. a N is the total number of randomized patients, patients that have been randomized more than once are only included according to their first randomization. Patients that are randomized but will not be included in the primary analysis are not included in this table; b significant stenosis on coronary angiography but no revascularization; c Defined as arterial obstructive disease involving at least 2 vascular beds where vascular bed involvement is characterized by either 1) CAD (defined as CAD, PCI or CABG), 2) PAD, 3) carotid artery stenosis or cerebral revascularization; d Defined as at least one of retinopathy, autonomic neuropathy, peripheral neuropathy, and nephropathy; e Medications used within 30 days of randomization, aspirin use is captured on day of randomization.
dose in THEMIS should impact efficacy less but could overestimate the overall bleeding risk, which is a conservative approach. [37][38][39] Multiple subgroup analyses have been prespecified, such as patients with multivessel coronary artery disease and prior percutaneous or surgical revascularization, but despite its size, the trial may not be well powered for subgroup analysis. Some peripheral ischemic events may not have been captured, as adjudication of these endpoints were added partway through the trial, though the trial was still blinded when this occurred, thus any assessment should be unbiased.

| CONCLUSION
Approaches to reduce CV morbidity further in patients with DM with stable atherosclerosis are urgently needed. More intense antiplatelet therapy is a promising approach. The THEMIS trial is assessing whether DAPT with ticagrelor and low-dose aspirin provides a significant reduction in ischemic events with an acceptable increase in bleeding. If the trial is positive overall, it will change the treatment paradigm for patients with diabetes. If the trial is clearly negative overall and in all major subgroups, it will provide a clear risk level below which more intense antiplatelet therapy is not indicated. Thus, irrespective of the exact results, THEMIS should greatly refine our understanding of the role of DAPT in patients across the atherothrombotic spectrum.

ACKNOWLEDGMENTS
We would like to thank Jayne Prats PhD for assistance with editing (limited to collation of author comments and formatting), with funding from AstraZeneca.