Non-vitamin K Antagonist Oral Anticoagulants Beyond Atrial Fibrillation

What Did We Learn From COMPASS and COMMANDER-HF?

Ahsan A. Khan; Gregory Y.H. Lip


Eur Heart J. 2019;40(46):3754-3756. 

The introduction of the non-vitamin K antagonist oral anticoagulants (NOACs) has been a game changer in our management of many thromboembolic conditions. The relative efficacy, safety, and convenience (rapid onset/offset of action, fewer drug and food interactions, and relatively predictable pharmacokinetics) of NOACs compared with oral vitamin K antagonists (VKAs, e.g. warfarin) has accelerated their use globally. Due to appreciable advantages over VKAs, the use of NOACs has been investigated in other medical conditions beyond stroke prevention in atrial fibrillation (AF) and the prevention or treatment of venous thromboembolism (VTE).[1,2]

Attention of the cardiovascular community has been directed to assess utility of NOACs in secondary prevention of coronary artery disease (CAD) and heart failure, two important conditions with life-limiting potential. Despite use of effective secondary prevention strategies, 5–10% of patients with known CAD have recurrent events annually.[3] Aspirin has been the mainstay of treatment for secondary prevention in such patients, with a 19% lowered risk of major adverse cardiovascular event and a 9% lower risk of cardiovascular death than placebo.[4] However, long-term treatment with a VKA alone or in a combination with aspirin is found to be superior to aspirin for secondary prevention but has been associated with an increased risk of bleeding.[5] In the ATLAS ACS 2-TIMI 51 trial, the addition of low-dose rivaroxaban (a factor Xa inhibitor) to dual antiplatelet therapy in patients with a recent acute coronary syndrome significantly reduced mortality and recurrent ischaemic events including stent thrombosis [in patients undergoing percutaneous coronary intervention (PCI)] for a median of 1.1 years when compared with placebo; however, this was at the cost of an increased risk of major bleeding and intracranial haemorrhage but not the risk of fatal bleeding.[6,7]

Two recent trials merit further discussion. The Cardiovascular Outcomes for People using Anticoagulation Strategies (COMPASS) trial was a double-blind, randomized placebo-controlled trial involving 27 395 participants from 602 centres in 33 countries comparing rivaroxaban with or without aspirin, and aspirin alone.[8] The trial was stopped early after a mean follow-up of 23 months showing that in patients with stable CAD, low-dose rivaroxaban [2.5 milligrams (mg) twice daily (b.i.d.)] plus aspirin [100 mg once daily (o.d.)] resulted in better cardiovascular outcomes compared with aspirin alone.[8] Despite more non-fatal bleeding events, intracranial and fatal bleeding event rates were similar comparing the rivaroxaban + aspirin and aspirin only groups.[8] Interestingly, major bleeding was noted to be twice as high in the elderly subgroup (age ≥75 years) in the rivaroxaban + aspirin group compared with the aspirin group but this was not found to be statistically significant. The higher-dose rivaroxaban (5 mg b.i.d.) arm did not lead to better cardiovascular outcomes compared with aspirin, at the expense of a similar major bleeding event rate.[8]

Perhaps the more impressive results were seen in the large subgroup of patients with peripheral artery disease (PAD). The COMPASS-PAD substudy enrolled 7470 patients with PAD, showed that combination of low-dose rivaroxaban and aspirin reduced the composite endpoint of cardiovascular death, myocardial infarction (MI), or stroke and major adverse limb events including major amputation.[9] Again, there was an increase in major bleeding (mainly gastrointestinal) in this group compared with the aspirin group, but not fatal or critical organ bleeding.[9] Notably, a landmark post hoc analysis of COMPASS data explored the effects of treatment during Year 1, Year 2, and beyond Year 2 and discovered that the risk of bleeding from addition of rivaroxaban to aspirin decreased after Year 1, while the reduction in primary outcome (first occurrence of stroke, MI, or cardiovascular death) remains relatively constant throughout all time intervals.[8]

More recently, the use of NOACs has been evaluated in the setting of heart failure. Indeed, patients with heart failure and reduced left ventricular function are associated with a high morbidity and mortality, and thromboembolic complications are common. Of note, heart failure is typically associated with older age, as well as AF (often asymptomatic in elderly patients) and pathophysiologically, with a prothrombotic state. Thus, there was some rationale for considering the use of anticoagulation, although prior trials in the warfarin era did not reduce adverse outcomes but increased bleeding.[10,11] Perhaps, the NOACs may fare better.

The COMMANDER-HF trial was designed to assess the effectiveness and safety of rivaroxaban in reducing thrombin load, and therefore, mortality, MI, or stroke in patients with worsening chronic heart failure, reduced left ventricular ejection fraction, CAD, and no AF.[12] This was a double-blind, randomized controlled trial involving 5022 patients.[12] Patients were randomly assigned to receive rivaroxaban at a dose of 2.5 mg b.i.d. or placebo in addition to standard care following treatment for an episode of worsening heart failure.[12] Over a median follow-up of 21.1 months, there was no significant difference noted in death, MI, or stroke between the two groups, nor the primary safety outcome of fatal bleeding or bleeding into a critical space with a potential for causing permanent disability.[12]

So where does this leave us? The efficacy and safety of NOACs in the setting of AF and VTE are widely recognized. The COMPASS trial has shown that rivaroxaban does have a role in secondary prevention amongst high-risk patients with CAD and PAD, but requires appropriate patient selection and careful consideration of risk-benefit balance. In contrast, the COMMANDER-HF trial failed to show a significant difference in primary outcome between the rivaroxaban and placebo groups. This follows another large clinical trial (NAVIGATE-ESUS), where the use of rivaroxaban in patients with embolic stroke of undetermined source (ESUS) was not superior to aspirin with regard to the prevention of recurrent stroke and was associated with a 2.7-fold higher risk of bleeding that led to its early termination.[13] Unpublished results for another similar trial with dabigatran, RESPECT-ESUS, that were presented at the World Stroke Congress in Montreal, Canada (October 2018) also showed no significant reduction in recurrent stroke, but similar bleeding risks, between dabigatran and aspirin treated patients with ESUS.[14]

Where do we go from here? It is important to appreciate the many known unknowns in chronic cardiovascular conditions. For example, AF due to its heterogeneity can present in any individual at any time. Clearly, there are certain markers that point towards an increased risk of incident AF (which may be asymptomatic) such as structural heart disease, presence of heart failure, hypertension, increasing age, diabetes, and evidence of coronary or peripheral artery disease which in turn lead to an increased risk of stroke given that the risk factors for incident AF are often the same risk factors that increase stroke risk.[15,16] As majority of these risk factors are common in patients with underlying CAD or PAD, or in patients developing heart failure, the possibility of screening for silent AF in such patients, especially if they have presented with a stroke, should be considered.[17]

While there appears to be some role for NOACs in high-risk patients with CAD or PAD, adding an antiplatelet such as aspirin to an oral anticoagulant, irrespective of dose increases bleeding risk. Evidently, certain risk factors such as old age, poor renal function, high-risk patients (for example chronic liver disease, previous history of bleeding or those known to have a bleeding diathesis), and those taking concomitant medications (for example bisphosphonates) may all increase the risk of bleeding further, with an important interaction between modifiable and non-modifiable bleeding risk factors that leads to determination of bleeding events.[18] Of note, bleeding risk is dynamic, and often the change in bleeding risk profile is more strongly associated with bleeding events than a 'one off' bleeding risk assessment at baseline, emphasizing the necessity to re-assess bleeding risk at every patient contact.[19]

If NOACs are useful for patients at 'high risk' of thrombotic events, such patients are also at high risk for bleeding outcomes. Careful risk stratification, and frequent reassessment would be prudent, as well as clear delineation of patients suited for treatment. Perhaps another perspective is to ask whether aspirin is a drug whose time has gone, given that in the setting of CAD or PAD, better alternatives are evident as secondary prevention. Even in AF patients undergoing PCIs, recent trials have suggested that omission of aspirin in a dual therapy strategy may be a safer option compared with a triple therapy approach.[20] Ultimately, the use of anticoagulants such as NOACs requires a balance between the reduction in thromboembolism and the potential for harm by serious bleeding (Figure 1).

Figure 1.

The role of NOACs in a prothrombotic state compared to aspirin.