Study summaries — NERDCAT

Wallentin L, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndrome. N Engl J Med 2009;361:1045-57.

Bottom line: In patients with ACS (excluding STEMI treated with fibrinolysis, or low-risk unstable angina), ticagrelor reduced the risk of death (NNT 72) or MI (NNT 91) compared to clopidogrel at 1 year. Conversely, ticagrelor increased the risk of certain adverse events compared to clopidogrel, including non-CABG major bleed (NNH 143) and dyspnea (NNH 17).

Patients

Inclusion:
- ACS with symptom onset <24h, including:
  - STEMI treated with primary PCI
  - NSTE-ACS with at least 2 of the following:
    - EKG - ST-segment changes suggesting ischemia
    - Troponin or CK-MB positive
    - 1+ additional risk factor (age 60+ y; previous MI, CABG or ischemic stroke/TIA; CAD with stenosis at least 50% in 2+ vessels; carotid stenosis of at least 50% or cerebral revascularization; CKD with eGFR <60 mL/min; diabetes; PAD)
Exclusion:
- Fibrinolytic <24h before randomization
- Need for oral anticoagulation
- Increased risk of bradycardia
- Taking medication that is a strong CYP 3A inhibitor/inducer
? screened -> 18,624 randomized & analyzed for efficacy (18,421 analyzed for safety outcomes)
"Average" patient
- Age 62 y (75+ y - 15%)
- Female 28%
- White 92%
- ACS final diagnosis
  - STEMI 38%
  - NSTEMI 42%
  - Unstable angina 17%
- CV history
  - MI 21%
  - PCI 13%
  - CABG 6%
  - HF 5%
  - Non-hemorrhagic stroke 4%
  - PAD 6%
- CV risk factors
  - Smoker 36%
  - HTN 65%
  - CKD 4%
  - Dyslipidemia 47%
  - Diabetes 25%

Interventions & co-interventions

I: Ticagrelor x up to 12 months (median 9 months)
- Loading dose of 180 mg PO x1, followed by
- Maintenance dose of 90 mg PO BID
- Started at median 11 hours from start of chest pain
- % of patients taking at least 80% of study drug: 83%
C: Clopidogrel x up to 12 months (median 9 months)
- Loading dose of 300 mg PO x1 (additional 300 mg for total loading dose 600 mg could be given prior to PCI), followed by
- Maintenance dose of 75 mg PO once daily
- Started at median 11 hours from start of chest pain
- % of patients taking at least 80% of study drug: 83%
Co-interventions
- ASA 75-100 mg PO daily (could be increased to 325 mg PO daily for 1st 6 months after stent placement)
- Procedures
  - Coronary angiography 81%
  - PCI during index hospitalization 61%
- Other meds at discharge
  - Statin, ACEI/ARB, beta-blocker ~90%
  - PPI 45%

Results @ median 9 months

Statistically significant reduction with ticagrelor in:
- Death from any cause: 4.5% vs 5.9% (NNT 72)
- Primary outcome (vascular death, MI, stroke): Ticagrelor 9.8%, clopidogrel 11.7% (NNT 53)
- MI 5.8% vs 6.9% (NNT 91)
- Stent thrombosis (probable or definite): 2.2% vs 2.9% (NNT 143)
No statistically significant difference in:
- Ischemic stroke: 1.1% in both groups
- Hemorrhagic stroke: 0.2% vs 0.1% (p=0.10)
- Recurrent ischemia: 5.8% vs 6.2% (p=0.22)
Safety:
- Premature discontinuation: 23.4% vs 21.5% (NNH 53)
  - Because of adverse event: 7.4% vs 6.0% (NNH 72)
- Major bleed: 11.6% vs 11.2% (p=0.43)
  - Not related to CABG: 4.5% vs 3.8% (NNH 143)
- Dyspnea: 13.8% vs 7.8% (NNH 17)
  - Requiring study drug discontinuation: 0.9% vs 0.1% (NNH 125)
- Bradycardia requiring pacemaker insertion: 0.9% in both groups
- Transient reversible increases in SCr & uric acid greater in ticagrelor group
Subgroup analyses: 3 of 25 subgroups had a positive test for interaction (each with p=0.05 or lower). Of interest:
- Region: Patients in the US did not have a reduction in the primary outcome (HR 1.25, 0.93-1.67), whereas all other groups did, including Africa, Asia, Australia, Europe, and Middle East. Further analysis of this subgroup effect suggested that this may have resulted from American sites preferentially using ASA doses of 300 mg/day or greater, whereas other sites preferentially used 75-100 mg/d.
- Lipid-lowering therapy: Patients also receiving lipid-lowering therapy benefited more than those not receiving lipid-lowering therapy. This was not further analyzed, but is likely due to chance or confounding by unmeasured factors.

Issues with internal validity?

Unclear risk of bias: Incompletely described as a "randomized, double-blind trial". No description of sequence generation, allocation concealment, blinding method.
Low risk of attrition bias (loss-to-follow-up <0.5%), and analysis properly included all randomized patients (intention-to-treat population).

Bhatt DL, et al. Clopidogrel with or without omeprazole in coronary artery disease. N Engl J Med 2010;363:1909-17

Bottom Line: In patients on clopidogrel + ASA following ACS or stent placement, the addition of omeprazole reduced the risk of clinical GI events (NNT 56 over 6 months), including overt upper GI bleed (NNT 100 over 6 months) with no increase in CV events. Because this trial was stopped early, the benefits (& reported NNTs) may be overestimated.

Integrating This Study Into Practice

Despite this uncertainty caused from stopping the trial early, it's reasonable to consider a PPI for the duration of dual antiplatelet therapy in patients without prior GI bleed, especially if they have other risk factors or medications that further increase their risk. PPIs should be mandatory in those with prior GI bleed while taking dual antiplatelet therapy unless contraindicated.

Design

Allocation-concealed "superiority" RCT with all (patients, clinicians, outcome adjudicators) blinded, low loss-to-follow-up (3%), analyzed using the intention-to-treat population.

Study stopped early due to abrupt loss of funding (sponsor went bankrupt), leading to accumulation of only ~1/3 of planned GI primary outcome events.

Patients and Setting

15 countries, 393 centers
Enrolled between January-December 2008
Inclusion criteria:
1. Age 21+ y/o
  - (designed to have >60% of patients 65+ y/o)
2. Required DAPT (ASA + clopidogrel) x 12+ months, specifically
  - NSTE-ACS
  - STEMI
  - Coronary stent implantation
Key exclusion criteria:
- Erosive esophagitis, esophageal or gastric varices, non-endoscopic gastric surgery
- Hx of hemorrhagic stroke, intracranial neoplasm, AVM, aneurysm
- Active pathological bleeding or hx of hereditary/acquired hemostatic disorder
- Other indication for gastroprotection (PPI, H2RA, sucralfate, misoprostol)
- CABG <30 days prior to randomization
- Cardiogenic shock, refractory ventricular arrhythmias or HF with NYHA class IV at time of randomization
- Meds
  - >21 days of clopidogrel or another thienopyridine prior to randomization
  - Needs oral anticoagulation
  - Recent fibrinolytic therapy
  - Steroids equivalent to >5 mg/d of prednisone
- Labs
  - Hemoglobin <100 g/L
  - Platelets <100
4444 screened for eligibility -> 3873 randomized -> 3761 analyzed
Average patient:
- 66 y/o
- Female 32%
- Race: White 94%
- BMI 28
- CV hx
  - PCI 72%
  - ACS 42%
  - MI 30%
  - PAD 12%
  - Stroke 8%
  - Other vascular disease 50%
- Hx of GI bleed/ulcer 4%
- PMHx
  - Current smoker 13%
    - HTN 80%
    - Diabetes 30%
    - Dyslipidemia 78%

Intervention and Control

Intervention: Omeprazole 20 mg/d
- Note: Part of proprietary product CGT-2168, which is a fixed-dose combination of clopidogrel 75 mg + omeprazole 20 mg
Control: Placebo
Co-interventions common to both groups:
- ASA 75-325 mg/d + clopidogrel 75 mg/d (planned duration of at least 12 months)

Outcomes

Median follow-up 106 days (max 341 days)
- Note: Outcomes standardized to 180 days
Primary efficacy outcome: Composite upper GI events (ulcer complications [perforation, obstruction, bleed], occult bleeds presumed to be from GI, symptomatic non-bleeding ulcer or erosions
Primary safety outcome: Composite CV events (CV death/ischemic stroke/non-fatal MI/coronary revascularization)
Subgroup analysis: Outcomes did not differ based on baseline H pylori status or use/non-use of non-ASA NSAIDs (randomization stratified based on these 2 factors)

Key Considerations & Interpretation

Trial stopped early: This trial with otherwise low risk of bias was truncated (stopped early) due to loss of funding. Truncated trials tend to exaggerate effect sizes (overestimate benefits/harms), especially when few events have occurred. It's therefore possible that the observed 1.8% absolute risk reduction (NNT 56) in clinical GI events over 6 months seen with omeprazole versus clopidogrel in COGENT based on only 55 GI events may be an overestimate of the real benefit.

Who does this apply to?

COGENT enrolled patients with new indication for 12 months of DAPT (3/4 PCI, 1/4 ACS), without high-risk features for GI bleed. This is therefore a trial to determine whether routine use of PPIs with DAPT is beneficial in patients at otherwise low risk of GI bleed. Given the issue with stopping early described above, the GI benefits of routinely adding a PPI to DAPT in patients without previous GI bleed/ulcer remain uncertain.
This doesn't apply to patients with prior upper GI bleed, where previous trials have demonstrated a large reduction in GI events when adding a PPI to antiplatelet therapy in patients who've previously bled (1, 2).

Does this trial prove that PPIs don't increase CV risk?

COGENT wasn't designed to demonstrate the absolute CV safety of omeprazole (or to dispel the possible interaction with clopidogrel), and in isolation can't be used to demonstrate safety. Despite this, omeprazole did not numerically increase the risk of CV events in this population at high risk of CV events (4.9% versus 5.7% with placebo).
Interpreted in the context of all of the observational and pharmacokinetic evidence, it appears that the association between PPIs & increased CV events may be due to confounding factors (i.e. weight, smoking, and other lifestyle factors that predispose a patient to be prescribed a PPI) rather than a true cause-and-effect.

PLATO - Ticagrelor in ACS