REDUCE-AMI: Beta-blockers in MI & LVEF >=50%

REDUCE-AMI: NEJM 2024 DOI: 10.1056/NEJMoa2401479

Bottom line: In patients with myocardial infarction, LVEF >=50%, and no other indication for beta-blockers (angina, arrhythmia, heart failure), beta-blockers do not meaningfully reduce death or recurrent myocardial infarction. Additional trials that should complete in 2024 will provide additional data, including in patients with LVEF 41-49%.

Patients (n=5020 randomized)

• Setting: Sweden (95%), Estonia, NZ, 2017-2023

• Included:

  • Adults 1-7 days from MI

  • Obstructive coronary artery disease (CAD) on coronary angiography (i.e. stenosis >=50%, FFR <=0.8 or iFR <=0.89 in any segment)

  • Echo after MI showing left ventricular ejection fraction (LVEF) >=50% (i.e. preserved)

• Excluded: Any other indication for beta-blockers; any contraindication to beta-blockers.

• Baseline:

  • Age 65, 22% female

  • Median 2 days from admission to randomization

  • LVEF & coronary angiography variables not reported

  • STEMI 35%, PCI 95-96%

  • Comorbidities: Prior MI ~7%, diabetes 14%, HTN 46%, smoking ~20%

  • Medications:

    • Beta-blocker prior to admission ~11-12%

    • At discharge: ASA/P2Y12 inhibitor 95-98%, ACEI/ARB 80%, statin, 98-99%

Intervention: Beta-blocker

• Metoprolol (1st choice; target >=100 mg/d) or bisoprolol (target >=5 mg/d)

• 96% prescribed beta-blocker on discharge (2/3 metoprolol, 1/3 bisoprolol), ~91% receiving at “month 2” visit, 82% receiving at “1-year” visit

Comparator: No beta-blocker

• 10% prescribed beta-blocker on discharge

Outcomes @ median 3.5 years

Primary outcome (death from any cause or new MI): Beta-blocker 7.9%, no beta-blocker 8.3% (hazard ratio [HR] 0.96, 95% confidence interval [CI] 0.79-1.16)

• Death: 3.9% vs 4.1% (HR 0.94, 95% CI 0.71-1.24)

• MI: 4.5% vs 4.7%

• Bayesian re-analysis (using https://benjamin-andrew.shinyapps.io/bayesian_trials/): Assuming non-informative prior (ignoring outdated post-MI beta-blocker RCTs from the pre-reperfusion era), posterior probability=19% of an absolute risk reduction of >=1% absolute risk reduction (HR<=0.88) at 3.5 years

Safety

• Hospitalization for bradycardia, 2-30 AVB, hypotension, syncope, or PPM implantation: 3.4% vs 3.2% (HR 1.08, 95% CI 0.79-1.46)

• Hospitalization for asthma or COPD: 0.6% in both groups (HR 0.94, 0.46-1.89)

Internal validity

• Randomization using permuted blocks

• Allocation via web-based system

• Performance bias: Moderate crossover (18% from beta-blocker to no beta-blocker by 1 year in intervention group; ~10% to beta-blocker in comparator group) biases the results toward the null

• Detection bias: No blinding of participants or treating clinicians (open-label), but objective (death from any cause) & semi-objective outcomes (e.g. MI hospitalization) minimize risk of detection bias

• No loss to follow-up (but missing data for 8 who emigrated, 4 withdrew consent)

• Intention-to-treat analysis

Generalizability & other considerations

• Who does these results NOT apply to?

  • These results do NOT apply to patients with another valid cardiovascular indication for beta-blockers, including:

    • Angina (despite PCI/CABG, if indicated)

    • Arrhythmia (ventricular arrhythmia, atrial arrhythmia requiring rate control, congenital long QT syndrome, etc)

    • Ejection fraction <50% [especially those with heart failure]

    • HF with reduced/mildly-reduced/improved LVEF

• With that said, these results broadly apply to patients not captured within the above, which is most patients with MI seen in contemporary practice

• This is the first contemporary trial of beta-blockers post-MI (see https://nerdcat.org/studysummaries/beta-blockers-cad for our summary of this previous evidence)

  • Prior to this trial, practice was mainly driven by a 1999 meta-analysis of RCTs conducted 1966-1991; prior to widespread use of now-established treatments for ACS/MI (especially reperfusion with PCI/fibrinolytics, ASA, statins)

• Several other ongoing trials will shed further light on the role of beta-blockers in patients post-MI with LVEF >40%, including:

  • ABYSS (n=3700, France): Beta-blocker continuation vs discontinuation 6 months post-MI (excluding patients with LVEF <40%, persistent angina/ischemia, HF in last 2 y, arrhythmia)

  • BETAMI (n=2900, Norway) & DANBLOCK (n=2760, Denmark): Beta-blocker vs no beta-blocker post-MI (excluding patients with LVEF <40%, clinical heart failure, LV akinesia in >=3 segments, arrhythmia)

POPular Genetics: Using pharmacogenomics to guide antiplatelet management

Claassens DMF, et al. A genotype-guided strategy for oral P2Y12 inhibitors in primary PCI. NEJM 2019;381:1621-31.

Bottom line: Among STEMI patients undergoing primary PCI, CYP2C19 genotype-guided P2Y12 inhibitor selection (leading to targeted de-escalation to clopidogrel in 2/3 of patients) reduced the risk of minor bleeding by ~3% without increase thrombotic events over 1 year.

Patients (n=2751 randomized, n=2488 analyzed)

• Included:

  • STEMI treated with primary PCI with stent

• Excluded

  • Severe HTN (>180/110 mm Hg)

  • Cardiogenic shock (SBP ≤80 mm Hg for >30 min)

  • Active malignancy causing increased bleed risk (investigator’s opinion)

  • Dialysis-dependent CKD

• Baseline

  • Age 61, female 25%

  • Prior coronary stent 8%, prior MI 7-8%

  • Prior bleed 2%

  • Current smoker 46%, HTN 42%, diabetes 11%, CrCl <60 9%

  • Treated vessel: LAD 43%, RCA 42%, bifurcation lesion ~20%

  • Drug-eluting stent in 94%

  • Total stent length 28 mm

Intervention: CYP2C19 genotype-guided P2Y12 inhibitor de-escalation x12 months

• Assay: TaqMan StepOnePlus assay (central lab) or Spartan RX (point-of-care test) ASAP after randomization

• Tested for CYP2C19*2 and CYP2C19*3 loss-of-function alleles

• Genotype:

  • Extensive metabolizer (good clopidogrel response; *1/*1): 67%

  • Intermediate metabolizer (*1/*2 or *1/*3): 29%

  • Poor metabolizer (*2/*2, *2/*3 or *3/*3): 2-3%

  • Not done: 1.4%

• Strategy:

  • If any *2 or *3: Ticagrelor or prasugrel

  • Neither (extensive metabolizer): Clopidogrel

• Selected: Clopidogrel (61%), ticagrelor (38%), prasugrel (1%)

Comparator: Standard P2Y12 inhibitor selection x12 months

• Selected: ticagrelor (91%), clopidogrel (7%), prasugrel (2%)

Internal Validity: Low risk of allocation & attrition bias; unclear risk of performance & detection bias

• Computer-generated block randomization

• Internet-based allocation

• Open-label (patients, clinicians aware of allocation after randomization)

• Blinded adjudication committee

• Intention-to-treat (ITT) & per-protocol analyses (ITT for superiority, ITT+PP for non-inferiority)

• 3 patients lost to follow-up

Outcomes @ 12 months

• Composite

  • Definition: Death/MI/definite stent thrombosis/stroke/major bleeding (PLATO definition)

  • Genotype-guided: 5.1% vs control 5.9%

  • Hazard ratio (HR) 0.87 (95% confidence interval 0.62-1.21)

  • Absolute difference: -0.7% (-2.0% to 0.7%), meeting study’s non-inferiority criteria

• Death: 1.5% in both groups (HR 1.00, 0.53-1.89)

• MI: 1.5% vs 2.1% (HR 0.73, 0.41-1.32)

• Definite stent thrombosis: 0.2% in both groups (HR 0.67, 0.11-4.01)

• Stroke: 0.6% vs 0.9% (HR 0.73, 0.29-1.82)

• Major or minor bleed: 9.8% vs 12.5% (HR 0.78, 0.61-0.98)

  • Major bleed: 2.3% in both groups (HR 0.97, 0.58-1.63)

  • Minor bleed: 7.6% vs 10.5% (HR 0.72, 0.55-0.94)

• Dyspnea: Not compared between groups

Other Considerations

• Comparator changed from clopidogrel to ticagrelor/prasugrel part-way through the trial; patients enrolled before this amendment excluded from analyses

• Non-inferiority trial

  • Non-inferiority margin set as a 2% absolute risk increase for both primary outcomes (not well justified, large)

• Unlike PHARMCLO, POPular Genetics did not test for CYP2C19*17 (ultra-fast metabolizer; increased clopidogrel efficacy), or ABCB 13435

Other Studies

• Other studied de-escalation strategies include empiric de-escalation to clopidogrel after 1 month of potent P2Y12 inhibition, as well as platelet function testing-guided de-escalation

• PHARMCLO

  • Patients: 888 ACS patients in Europe

    • Baseline:

      • Age 71, female 32%

      • STEMI 27%, NSTEMI 68%, UA 2%, no ACS 3%

      • Prior PCI 19%, prior MI 21%

      • Current smoker 22%, HTN 74%, diabetes 26%, CKD 9%

      • 96% underwent coronary angiography: 62% got PCI, 11% CABG

      • Treated vessel: LAD 54%, RCA 47%

      • Genotype (intervention group)

        • ABCB1 3435 mutation: 47%

        • CYP2C19*2 29%, *2/*2 4%

        • CYP2C19*17 31%, *17/*17 8%

  • Intervention: Genotype-guided P2Y12 inhibitor selection (immediately on ACS diagnosis) x12 months

    • Assay: ST Q3 (point-of-care test that takes ~70 min for result)

    • Tested for ABCB13435, CYP2C19*2, and CYP2C19*17 (increased clopidogrel efficacy)

    • Selected: Clopidogrel 43%, ticagrelor 43%, prasugrel 8%

  • Control: Standard P2Y12 inhibitor selection x12 months

    • Selected: Clopidogrel 51%, ticagrelor 33%, prasugrel 8%

  • Outcomes:

    • Composite (CV death, MI, stroke, major bleed (BARC 3-5): 15.9% vs 25.9% (HR 0.58, 0.43-0.78)

    • CV death/MI/stroke: 12.9% vs 21.4% (HR 0.57, 0.41-0.80)

    • Major bleed: 4.2% vs 6.8% (HR 0.62, 0.35-1.11)

  • Caveats:

    • Stopped prematurely 1/4 of the way through by ethics committee as genotyping assay not previously certified; therefore, the observed benefit of genotype-guided intervention is likely a large overestimate

TWILIGHT: DAPT x3 months followed by ticagrelor monotherapy vs continued DAPT after PCI

Mehran R, et al. Ticagrelor with or without aspirin in high-risk patients with PCI. NEJM 2019;381:2032-42.

Bottom Line:

• Among patients at high risk of thrombosis who underwent PCI & tolerated ticagrelor + ASA for 3 months without thrombotic or major bleeding events, switching to ticagrelor monotherapy reduced the risk of bleeding (NNT=34) with similar risk of thrombotic events vs continued ticagrelor+ASA for up to 15 months.

• It remains unclear how this “TWILIGHT” regimen compares to a standard regimen of ASA + clopidogrel, or other novel antiplatelet de-escalation strategies), in the era of 2nd-generation drug-eluting stents with low risk of stent thrombosis.

Patients (n=7119)

• 9006 enrolled at time of PCI -> 7119 randomized 3 months post-PCI

  • 21% enrolled excluded before randomization, mostly due to non-adherence (13% of enrolled, 1/2 of related to dyspnea) or major thrombotic/bleeding event (3% of enrolled) in 1st 3 months post-PCI

• Inclusion:

  • Underwent successful PCI with drug-eluting stent, with “intent to discharge the patient with a regimen of ticagrelor + ASA”

  • At least 1 clinical feature + 1 angiographic feature associated with high thrombotic or bleeding risk:

    • Clinical: Age >65 y, female, troponin+ ACS, established vascular disease (previous MI, known PAD or CAD/PAD revascularization), diabetes treated with medications, CKD

    • Angiographic: Multivessel CAD, total stent length >30 mm, thrombotic target lesion, bifurcation lesion treated with 2 stents, obstructive LM or pLAD lesion, or calcified target lesion treated with atherectomy

  • Did not experience major bleed (BARC 3b, 4 or 5) or thrombotic event (stroke, MI, coronary revascularization) in 1st 3 months post-PCI

  • Adherent to ticagrelor + ASA based on manual pill count after 1st 3 months

• Excluded:

  • Presentation: STEMI, fibrinolytic therapy <24h of PCI, cardiogenic shock

  • PMHx: Prior stroke, dialysis-dependent CKD, cirrhosis, life expectancy <1 y contraindication to ASA/ticagrelor

  • Other: Ongoing OAC use, strong CYP3A inducer/inhibitor, active bleeding or “extreme” risk for major bleed, platelet count <100

• Baseline

  • Age 65, female 24%, non-white 31%, North America 42%

  • Indication for PCI: NSTEMI 30%, UA 35%, stable angina 29%, asymptomatic 6%

  • Multivessel CAD 63%, previous MI 29%, previous PCI 42%, previous CABG 10%, PAD 7%

  • Smoker 22%, diabetes 37%, CKD 17%

  • Anemia 20%, previous major bleed 1%

  • Angiographic/procedural:

    • LM treated in ~5%, LAD in 56%

    • Mean # of vessels treated 1.3, mean total stent length 40 mm, mean minimum stent diameter 2.8 mm

    • Bifurcation 12%, thrombus ~10%, chronic total occlusion 6%, SVG 2%

    • 2nd-generation drug-eluting stent used in >97%

Interventions & Controls

• For the 1st 3 months after PCI, all patients received open-label enteric-coated ASA 81-100 mg/d + ticagrelor 90 mg BID

• After the 1st 3 months, randomized to open-label ticagrelor plus 12 months of:

• Intervention (ticagrelor monotherapy after 3 months of ticagrelor-based DAPT): Placebo

• Control (continued ticagrelor-based DAPT): ASA

• Similar % of patients took >80% of ticagrelor (86-87%) & study drug (82-83%)

Outcomes @ month 3 to 15 (during study intervention after 1st 3 months of DAPT)

No difference in thrombotic events

• Death/MI/stroke (per-protocol analysis): 3.9% in both groups

  • Hazard ratio (HR) 0.99 (95% confidence interval [CI] 0.78-1.25)

  • Absolute difference -0.1 (-1% to +0.8), meeting non-inferiority criteria (see below)

  • Death: 1.0% vs 1.3%

  • MI: 2.7% in both groups

  • Stroke: 0.5% vs 0.2%

• Stent thrombosis (definite/probable): 0.4% vs 0.6% (HR 0.74, 0.37-1.47)

Fewer bleeding events with abbreviated DAPT

• BARC 2, 3 or 5 bleed (1o outcome) @ month 15 after PCI: 4.0% vs 7.1% (NNT 34)

  • HR 0.56 (0.45-0.68)

  • Identical results for TIMI minor/major definition

  • BARC 3 or 5: 1% vs 2% (HR 0.49, 0.33-0.74)

• ISTH major: 1.1% vs 2.1% (HR 0.54, 0.37-0.80)

• GUSTO moderate/severe: 0.7% vs 1.4% (HR 0.53, 0.33-0.85)

Subgroup analyses: No subgroup effect for bleeding or thrombotic outcomes.

Internal Validity

• Low risk of

  • Allocation bias: Random sequence using permuted blocks stratified by site generated by independent statistician; concealed allocation via secure web-based system

  • Performance bias: Use of matching placebo (patients, clinicians, etc) unaware of treatment assignment

  • Detection bias: Outcomes adjudicated by committee unaware of study drug assignment

• Unclear risk of attrition bias

  • Loss-to-follow-up 1.6% for the primary outcome & 0.3% for death

  • Bleeding outcome analyzed using intention-to-treat principle (low risk), but thrombotic outcome analyzed using per-protocol (unclear risk)

Other Considerations

• Non-inferiority design for the key thrombotic endpoint (death/MI/stroke)

  • Pre-specified non-inferiority margin: Absolute difference of 1.6% assuming 8.0% incidence in the continued DAPT group

    • However, event rates lower than anticipated (3.9% instead of 8.0%). Ideally, a more conservative relative risk non-inferiority margin equivalent to a 1.6% increase above 8.0% would have been used (RR=9.6%/8.0%=1.20)

    • Using this stricter relative risk-based non-inferiority margin of 1.20, the study does not technically meet the study’s pre-specified non-inferiority criteria for this outcome.

  • Non-inferiority should only be concluded if both ITT & PP analyses consistently demonstrate non-inferiority; however, only PP analysis of thrombotic outcome reported in this trial.

Other Studies

• PLATO: The landmark study demonstrating the superiority of ticagrelor + ASA vs clopidogrel + ASA among all-comer ACS patients (except STEMI patients treated with fibrinolytics) in the era of bare-metal & 1st-generation drug-eluting stents.

• GLOBAL LEADERS: In a trial of patients who underwent PCI, ticagrelor-based DAPT x1 month, followed by ticagrelor monotherapy x23 months vs standard DAPT for ACS/stable angina did not reduce the risk of death/Q-wave MI.

• In a contemporary population-based registry study of ACS patients who underwent PCI, ticagrelor + ASA was not associated with a lower risk of major adverse coronary events compared to clopidogrel + ASA. However, it was associated with significantly more major bleeding & dyspnea

  • Similar to TWILIGHT, this cohort study was done in the era of second-generation drug-eluting stents with low risk of stent thrombosis.

ISAR-REACT 5: Ticagrelor vs prasugrel in ACS

Schupke S, et al. Ticagrelor or prasugrel in patients with acute coronary syndromes. NEJM 2019

Bottom line: In patients with ACS planned for invasive management (90% treated with new-gen drug-eluting stent), prasugrel reduced death/MI/stroke versus ticagrelor (NNT=42) at 1 year, which was driven by fewer MIs.

Patients (n=4018 randomized)

• Enrolled from 21 centres in Germany & 2 centres in Italy

• Included if:

  • Hospitalized for ACS (STEMI, NSTEMI or unstable angina)

  • Planned invasive strategy (i.e. scheduled for coronary angiogram)

• Key exclusion criteria:

  • Hx of any stroke, TIA or intracranial hemorrhage

  • Intracranial abnormality at risk of bleeding

  • Lysis <24h before randomization

  • Ticagrelor or prasugrel <5 days before randomization

• Average baseline characteristics

  • Age 65 y, female 24%

  • Diagnosis @ admission: STEMI 41%, NSTEMI 46%, UA 13%

  • Final diagnosis of ACS at discharge 91%

  • Index ACS managed with: PCI 84% (90% of which were drug-eluting stents), CABG 2%, medical management only 14%

  • Prior MI 16%, prior PCI 23%, prior CABG 6%

  • Cardiac risk factors: HTN 70%, diabetes 22%, dyslipidemia 58%

  • BMI 28, wt <60 kg 5%

Interventions: Ticagrelor ASAP vs Prasugrel x1 year

• Ticagrelor: Loading dose of 180 mg x1 ASAP after randomization (i.e. pre-loading before coronary angiogram), then 90 mg PO BID

• Prasugrel: Loading dose of 60 mg x1, then 10 mg once daily

  • If STEMI: Loading dose given ASAP after randomization

  • If NSTEMI/UA: Loading dose given after coronary angiogram

  • Lower maintenance dose of 5 mg daily given if age age 75+ or weight <60 kg

• ~20% in both groups were not discharged on their assigned study P2Y12i (most because they were not confirmed to have obstructive CAD on angiography, had indication for oral anticoagulant, or underwent CABG)

• Note: Patients did not get the medication for free as part of the trial; they had to obtain it as they otherwise would need to do in standard practice

Outcomes @ 1 year

• Primary outcome (composite of all-cause death, MI, stroke): Ticagrelor 9.3% vs prasugrel 6.9%

  • Hazard ratio (HR) 1.36, 95% confidence interval (CI) 1.09-1.70

  • NNT (using prasugrel instead of ticagrelor)=42 @ 1 year

• Secondary outcomes

  • Death: 4.5% vs 3.7% (HR 1.23, 95% CI 0.91-1.68)

  • CV death: 3.2% vs 3.0%

  • MI: 4.8% vs 3.0% (HR 1.63, 1.18-2.25), NNT=56 in favor of prasugrel

  • Definite/probable stent thrombosis: 1.3% vs 1.0%

  • Stroke: 1.1% vs 1.0%

  • Major bleed (BARC 3, 4 or 5): 5.4% vs 4.8% (HR 1.12, 0.83-1.51)

• Discontinued study intervention: Ticagrelor 15.2% vs prasugrel 12.5%

  • Median time to D/C: 84 vs 109 days

View fullsize

Risk of bias: Overall some concern

• Some concerns re: bias arising from the randomization process (allocation bias)

  • Computer-generated random sequence

  • Allocation concealment only by sealed, opaque envelopes (no details on storage location, sequential numbering or assignment); prone to tampering & less robust than other methods

• Some concerns re: bias due to deviations from intended interventions

  • Open-label design

  • Higher & earlier discontinuation of study drug in ticagrelor group, though no other evidence of differences in management between groups

  • Predicted direction of bias in favor of prasugrel for the primary efficacy outcome

• Low risk of bias due to missing outcome data (attrition bias)

  • Low loss-to-follow-up: <1% in both groups

  • Analyzed the intention-to-treat population

• Low risk of bias in measurement of the outcome (detection bias)

  • Open-label design; however objective outcome definition (especially for MI, the driver of the difference between ticagrelor & prasugrel), & outcomes adjudicators not aware of assigned intervention

• Low risk of bias in selection of the reported result (outcome reporting bias)

Other considerations

• The findings of this trial conflict with results of prior trials of prasugrel & ticagrelor:

  • The results of the primary outcome of ISAR-REACT 5 expressed as prasugrel vs ticagrelor are 6.9% vs 9.3%, HR 0.74 (0.59-0.92)

  • In the PLATO trial comparing ticagrelor to clopidogrel in ACS patients managed medically or with PCI, ticagrelor reduced death/MI/stroke vs clopidogrel.

    • 10.2% vs 12.3%, HR 0.84 (0.77-0.92)

    • Notably, ticagrelor also reduced death vs clopidogrel (HR 0.78)

  • Similarly, in the TRITON-TIMI 38 trial comparing prasugrel to clopidogrel in ACS patients managed almost exclusively with PCI, prasugrel reduced death/MI/stroke vs clopidogrel @ 15 months

    • 10.7% vs 12.7%, HR 0.83 (0.75-0.92)

    • Prasugrel did not reduce death vs clopidogrel (HR 0.96)

  • Conversely, in TRILOGY, in which only medically-managed ACS patients were randomized, prasugrel did not reduce CV death/MI/stroke vs clopidogrel, HR 0.96 (0.86-1.07)

  • Note: The previous PRAGUE-18 trial (n=1230) that also compared prasugrel to ticagrelor was severely underpowered & inconclusive.

• In terms of pharmacodynamics, ticagrelor inhibits platelet function at least as well as prasugrel, irrespective of assay technique (e.g. Rollini et al., SWAP-2, SWAP-3).

• Adherence to the P2Y12 inhibitor was apparently assessed (as briefly described in the protocol), but not reported in sufficient detail in this article. Of patients who discontinued the study P2Y12 inhibitor, those in the ticagrelor discontinued approximately 1 month earlier than prasugrel, which may explain at least in part the early higher MI rate with ticagrelor vs prasugrel.

• Overall, it’s unclear how much the findings of this trial are attributable to:

  • A yet-unidentified pharmacodynamic benefit of prasugrel;

  • Lower adherence to ticagrelor due to BID dosing;

  • Greater & earlier discontinuation of ticagrelor due to dyspnea.

• Arguably, it’s also unclear how the efficacy of prasugrel compares to clopidogrel in the era of modern drug-eluting stents with incredibly low rates of stent thrombosis, and in this emerging era of complete revascularization.

AUGUSTUS - Antithrombotic regimens including apixaban vs warfarin, & aspirin vs placebo, in patients with AFib plus PCI &/or ACS

Reference: Lopes RD, et al. Antithrombotic therapy after acute coronary syndrome or PCI in atrial fibrillation. NEJM 2019.

Bottom line: In patients with atrial fibrillation who either undergo PCI and/or have ACS, in combination with a P2Y12 inhibitor (almost always clopidogrel):

• Apixaban reduces the risk of major or clinically-relevant non-major bleeding (NNT=24), hospitalizations (NNT=27), & stroke (NNT=84) compared to warfarin at 6 months;

• Aspirin (beyond the first week) increases the risk major or clinically-relevant non-major bleeding (NNH=15), without a clear effect on hospitalization/death or ischemic events compared to placebo at 6 months;

• Therefore, an antithrombotic regimen of apixaban + clopidogrel (without aspirin) should be routinely considered in these patients. Warfarin should be limited to patients for whom a DOAC is contraindicated, intolerable or unaffordable; & aspirin beyond the first week should be limited to patients with very high risk of stent thrombosis/recurrent coronary events.
  

Patients (n=4614 from 33 countries)

• Included if (all of the following):

  • Age 18+ years

  • Known AF (paroxysmal, persistent or permanent) with planned long-term oral anticoagulation

  • Recent (<14 days) ACS &/or PCI with plan for 6+ months of P2Y12 inhibitor

• Key exclusion criteria:

  • Other indication for anticoagulation (prosthetic valve, VTE, mitral stenosis, etc)

  • History of intracranial hemorrhage, ongoing bleeding or coagulopathy

  • Recent/planned CABG

  • “Severe” renal insufficiency

• Average baseline characteristics:

  • Age 71 years, male (71%), white (92%)

  • Qualifying event: ACS+PCI (37%), medically-managed ACS (24%), elective PCI (39%)

    • ~6.6 days from ACS/PCI to randomization

  • CHA2DS2-VASc ~4, HAS-BLED ~3

  • Prior stroke/TIA/thromboembolism (14%), HF (43%), HTN (88%), diabetes (36%)

  • SCr >133 (8%)

  • Previous oral anticoagulant (49%)

Interventions x6 months

• 2x2 factorial design: Patients were simultaneously randomized to apixaban vs warfarin & aspirin vs placebo within 14 days of ACS &/or PCI, so total of 4 different intervention groups.

• Management prior to randomization: At the discretion of treating physicians according to local standard of care (likely that all at least received DAPT +/- anticoagulation leading up to randomization, though not recorded/reported)

• Anticoagulation: Apixaban vs warfarin

  • Apixaban arm: 5 mg PO BID

    • Reduced to 2.5 mg PO BID if 2 of the following: Age >80 years, wt <60 kg, SCr >133 umol/L

    • Discontinued study regimen prematurely: 13%

  • Warfarin to target INR 2.0-3.0

    • Median time in therapeutic range (TTR)=59%; INR<2.0 23% of the time, INR>3.0 3% of the time

    • Discontinued study regimen prematurely: 14%

• Antiplatelet: Aspirin 81 mg PO daily vs matching placebo

  • Discontinued study drug prematurely: 15-17%

• All: P2Y12 inhibitor left at the discretion of the treating clinicians (clopidogrel 93%, prasugrel 1%, ticagrelor 6%)

• After 6 months, anticoagulation & antiplatelets were managed according to local standard of care (i.e. not standardized for the trial)

Results @ 6 months

• Primary outcome: Major or clinically-relevant non-major bleeding, ISTH definition

• Key secondary outcomes: Composite of death or hospitalization; composite of death or ischemic events (stroke, MI, definite/probable stent thrombosis, or urgent revascularization).

View fullsize

Outcomes at 6 months of apixaban versus warfarin in combination with P2Y12 inhibitor +/- aspirin

View fullsize

Outcomes at 6 months of aspirin versus placebo in combination with P2Y12 inhibitor + apixaban or warfarin

Risk of bias

• Low risk of: Allocation bias (allocation concealed via interactive voice-response system), attrition bias (low [0.3%] loss to follow-up & analyzed by intention-to-treat), outcome reporting bias (all outcomes of interest defined & reported).

• Variable risk of performance/detection bias:

  • Apixaban vs warfarin comparison was open-label (i.e. patients & clinicians aware of treatment assignment):

    • All outcomes were adjudicated by a blinded clinical endpoint committee, therefore providing some protection against (but not eliminating) detection bias.

  • Aspirin vs placebo comparison was blinded (patients, clinicians, outcome adjudicators unaware of treatment assignment): Low risk of performance & detection bias.