Sandbox:JC ATLAS example

ATLAS ACS 2-TIMI 51

Mega JL, et al. "Rivaroxaban in Patients with Recent Acute Coronary Syndrome". The New England Journal of Medicine. 2012. 366(1):9-19. PubMed [1] NEJM [2]

Title	Rivaroxaban in Patients with Recent Acute Coronary Syndrome
Abbreviation	ATLAS ACS 2-TIMI 51
Description	Anti-Xa Therapy to Lower Cardiovascular Events in Addition to Standard Therapy in Subjects with Acute Coronary Syndrome-Thrombolysis in Myocardial Infarction 51
Subspeciality	Cardiology
Disease	Acute Coronary Disease
Other disease	Myocardial Infraction
Intervention 1	Rivaroxaban
Intervention 2	Placebo
Description 2	Rivaroxaban reduces cardiovascular death, MI, and stroke, but increases nonfatal bleeding
Design description	Rivaroxaban after ACS

Clinical Question

In patients with recent ACS, does the addition of rivaroxaban to standard dual antiplatelet therapy improve CV morbidity and mortality?

Bottom Line

In patients with recent ACS, the addition of rivaroxaban to standard dual antiplatelet therapy reduced the composite of CV mortality, recurrent MI, or stroke, but increased the risk of nonfatal bleeding.

Major Points

Two meta-analyses of RCTs investigating the addition of warfarin to aspirin among ACS patients suggested that combination therapy may reduce the risk of ischemic events. However, this was partially offset by an increased risk of bleeding compared to monotherapy.

ATLAS ACS-2, TIMI 51 (2012) randomized 15,342 patients with recent ACS to either twice-daily doses of either 2.5mg or 5mg of rivaroxaban (an oral factor Xa inhibitor) vs. placebo. About 50% of patients had STEMI; 25% had NSTEMI; and 25% had UA. 93% received dual antiplatelet therapy with aspirin and either clopidogrel or ticlopidine. At a mean follow-up of 13.1 months, both dosages of rivaroxaban decreased the rate of the primary composite outcome of CV mortality, MI, or stroke (8.9% vs. 10.7%; HR 0.84), as well as the rate of stent thrombosis. Low-dose rivaroxaban reduced CV mortality and all-cause mortality, but this was not observed with the higher dose. The benefits of rivaroxaban were offset by an increased rate of non-fatal bleeding, including ICH, but there was no increase in fatal bleeding. A 2013 subgroup analysis on STEMI-only patients published elsewhere demonstrated similar results. In May 2012, the FDA rejected rivaroxaban in ACS patients because of concerns regarding missing data; an appeal was again rejected in March 2013.

Of note, APPRAISE-2 (2011) randomized patients with ACS to apixaban (another oral factor Xa inhibitor) or placebo. However, this was terminated early secondary to increased rate of major bleeding in the apixaban arm without reductions in CV death, MI or ischemic stroke.

Guidelines

No guidelines

Design

• Multicenter, double-blind, parallel group, randomized, placebo-controlled trial
• N=15,342 patients with ACS in prior 7 days
  • Rivaroxaban 2.5mg PO BID (n=5,114)
  • Rivaroxaban 5mg PO BID (n=5,115)
  • Placebo (n=5,113)
• Setting: 766 centers in 44 countries
• Enrollment: 2008-2011
• Analysis: modified intention-to-treat, intention-to-treat
• Mean follow-up: 13.1 months
• Primary efficacy outcome: composite of CV mortality, MI or stroke
• Primary safety outcome: TIMI major bleeding not related to CABG

Population

Inclusion Criteria

• Age ≥18 years; if age <55, also diagnosis of either DM or previous MI
• Diagnosis of ACS

Exclusion Criteria

• Platelets <90,000/mm³ or hemoglobin <10 g/dL
• CrCl <30 mL/min
• GI bleed in the prior 12 months
• Prior ICH
• Prior ischemic stroke or TIA on both aspirin and thienopyridine

Baseline Characteristics

• Age: 61.7 years
• Male: 74.7%
• White: 73.5%
• Asian: 20.8%
• Weight: 78.0 kg
• HTN: 67.4%
• DM: 32.0%
• HLD: 48.5%
• Previous MI: 26.9%
• Index diagnosis:
  • STEMI: 50.4%
  • NSTEMI: 25.6%
  • UA: 24.0%
• PCI or CABG on enrollment: 60.4%
• Median time to randomization: 4.6 days

Baseline Medications

• ASA: 98.6%
• Thienopyridine: 92.8%
• Beta-blocker: 66.1%
• ACE inhibitor or ARB: 39.0%
• Statin: 83.5%
• CCB: 15.0%

Interventions

• ACS stabilized prior to enrollment
• Enrollment within 7 days after hospital admission
• Randomized in 1:1:1 fashion either to rivaroxaban 2.5 mg BID or rivaroxaban 5 mg BID or placebo
• Standard medical therapy included low-dose aspirin and thienopyridine (either clopidogrel or ticlodipine)
• Follow up at 4 weeks, 12 weeks, and every 12 weeks thereafter

Outcomes

All comparisons are rivaroxaban 2.5mg vs. rivaroxaban 5mg vs. placebo (2.5mg vs. placebo | 5mg vs. placebo).

Primary Outcomes

Composite of CV mortality, MI, or stroke (efficacy)

9.1% vs. 8.8% vs. 10.7% (HR 0.84; P=0.02 | HR 0.85; P=0.03)

TIMI major bleeding not associated with CABG (safety)

1.8% vs. 2.4% vs. 0.6% (HR 3.46; P<0.001 | HR 4.47; P<0.001)

Secondary Outcome

Composite of all-cause mortality, MI or stroke

9.3% vs. 9.1% vs. 11.0% (HR 0.83; P=0.02 | HR 0.84; P=0.02)

Additional Analyses

CV mortality

2.7% vs. 4.0% vs. 4.1% (HR 0.66; P=0.002 | HR 0.94; P=0.63)

All-cause mortality

2.9% vs. 4.4% vs. 4.5% (HR 0.68; P=0.002 | HR 0.95; P=0.66)

MI

6.1% vs. 4.9% vs. 6.6% (HR 0.90; P=0.27 | HR 0.79; P=0.02)

Any stroke

1.4% vs. 1.7% vs. 1.2% (HR 1.13; P=0.56 | HR 1.34; P=0.15)

Ischemic stroke

1.4% vs. 1.7% vs. 1.2% (HR 0.89; P=0.64 | HR 1.05; P=0.84)

Stent thrombosis

2.2% vs. 2.3% vs. 2.9% (HR 0.65; P=0.02 | HR 0.73; P=0.08)

"Good" follow-up

Complete: 85% vs. 84% vs. 85%

Death: 3% vs. 4% vs. 4%

Total: 88% vs. 87% vs. 89%

"Bad" follow-up

Consent withdrawn: 9% vs. 9% vs. 8%

Lost: 0.2% vs. 0.3% vs. 0.3%

Other: 3% vs. 4% vs. 3%

Total: 11% vs. 12% vs. 13%

Adverse Events

TIMI minor bleeding

0.9% vs. 1.6% vs. 0.5% (HR 1.62; P=0.09 | HR 2.52; P<0.001)

TIMI bleeding requiring medical attention

12.9% vs. 16.2% vs. 7.5% (HR 1.79; P<0.001 | HR 2.39; P<0.001)

ICH

0.4% vs. 0.7% vs. 0.2% (HR 2.83; P=0.04 | HR 3.74; P=0.005)

Fatal bleeding

0.1% vs. 0.4% vs. 0.2% (HR 0.67; P=0.45 | HR 1.72; P=0.20)

Subgroup Analysis

Reduction in primary efficacy outcome with rivaroxaban was consistent among subgroups except for patients with history of stroke or TIA.

Criticisms

• FDA declined approval of rivaroxaban for ACS in May 2012, citing:
  • High incomplete follow up (12%)
  • Vital status missing (9%)
  • Uncounted deaths
  • Different rates of outcomes between first and second halves of the trial are concerning for informative censoring
• Generalizability
• Rivaroxaban is not available at lower dose formulations.

Curiously, the publisher has not released letters to the editor pertaining to this article and relevant criticisms are therefore limited.

Funding

• Johnson & Johnson and Bayer Healthcare, developer and manufacturer of rivaroxaban, trade name Xarelto.
• Authors with multiple disclosures

Further Reading

• Concurrent editorial
• JACC STEMI Subgroup Analysis Article - Mega JL et al. Rivaroxaban in Patients Stabilized after a ST-Elevation Myocardial Infarction: Results from the ATLAS ACS 2-TIMI 51 Trial. Journal of the American College of Cardiology (2013), e-publishing only.